May 8, 1913] 



NATURE 



247 



are not due to mere imagination. My experiments 

 have been so far directed towards the question 

 as to how far differences of dynamical touch can 

 be made to produce effects that can be nun ed b;j 

 an ear not specially trained to observe them. The 

 apparatus used in these experiments consists of a 

 horizontal lever fixed in front of a piano-player 

 of the usual standard type; the short arm of this 

 lever is connected by a wire passing over pulleys, 

 or by some other connection, with the small 

 auxiliary bellows of the player, and acts directly 

 on it, the usual spring being removed. The lever 

 carries a sliding weight by which the collapsing 

 tension of the bellows can be regulated. When 

 the bellows collapses it closes a kind of throttle 

 valve in the mechanism, thus cutting the air off 



Fig. 1. — Apparatus (pr< 



lly protected) for 



and producing an action similar to that effected 

 by a short sharp impulse applied to the pianoforte 

 keys. By increasing the load, by shifting the 

 weight towards the end FF, a heavier sustained 

 pressure is produced. 



When the same passage is played first with the 

 weight at PP and then with it at FF, a noticeable 

 difference is observed. The loudness can be made 

 about the same in both cases by suitably altering 

 the pressure on the pedals, but in the former 

 position the result is a brilliant "metallic" effect 

 in which the treble notes stand out conspicuously, 

 while in the latter position soft, mellow tones are 

 produced in which the bass notes predominate. 



The differentiation between bass and treble parts 

 NO. 2271, VOL. 91] 



o) chords is not necessarily inconsistent with the 

 ingle-variable theory, since the hammers are of 

 different mass in different parts of the scale, and 

 therefore undergo different accelerations when the 

 same variable force is applied to them. This 

 dynamical differentiation is more satisfactory than 

 the popular mechanical arrangement for controlling 

 the two halves of the keyboard, as it involves no 

 hard and fast dividing line. 



In a passage involving chords it is impossible 

 tu separate the effects due to differences between 

 the notes of a chord from any possible differences 

 between the harmonics of the notes, and there- 

 fore it is necessary to choose a solo passage in 

 order to effect a decisive test. I have shown such 

 tests to a good many people; a few notice either 

 no differences or only very slight 

 differences. On the other hand, 

 it is very surprising to find how 

 many people notice conspicuous 

 differences, and those who are 

 most successful in detecting them 

 are often people with no ear for 

 music and no previous musical 

 training. In most cases I ask 

 them to describe what they ob- 

 serve without previously _ pre- 

 paring them. It is thus evident 

 that the dynamical effects depend- 

 ent on these differences of touch, 

 so far from being negligible, 

 must have a marked influence on 

 the impressions formed by a large 

 proportion of both the musical 

 and unmusical people who attend 

 a first-class pianoforte recital. 



Something more than merely 

 sliding a weight is necessary to 

 approximate to the rendering of a 

 good pianist, who can vary his 

 action on the keys from one note 

 to the other. To effect the same 

 result the lever must be controlled 

 by hand as well, being pressed 

 or jerked from above or below 

 practically in an unlimited variety 

 of ways. An almost infinitesimal 

 touch of the finger will often 

 ^a"piano. P ia)er. cause a particular note to ring 



out brilliantly. In the commercial 

 player the performer has to depend mainly for these 

 effects on his feet. Now not only are feet much 

 less sensitive than fingers, but the effects are 

 so modified by the elasticity of the various springs 

 that these latter have a predominating influence 

 in governing the touch, and a considerable effort 

 on die part of the performer often produces only 

 a slight difference in the result. 



A remarkable instance of how a trifling cause 

 may greatly modify pianoforte tones was shown 

 in a recent demonstration at the Physical Society, 

 when the effects were observed to be rather loud 

 and harsh. In this case a copper wire was used 

 for the connection between the lever and bellows 

 instead of one of steel. On trying the copper con- 



