342 



NATURE 



{July 25, 1878 



Note. — The analyses put within brackiH ari a sem'torie diffirjnt in 

 pitch from tha others in the grjup, and have their pitch marked at the side. 

 They have been taken where there did njt happen to be an example on th; 

 exact pitch wanted. 



On G voice i gave the following series : — 



1° I "3 o I "5 4"o o*8 0*4, 



On F another voice (No. 6) gave 



6" 2'2 i*o 1*5 o'8 3"4 o*i. 



In this last example we see that the fifth partial was much 

 greater than the prime. 



For the sake of brevity we have only given a few cases. Our 

 results contain the analyses of more than a hundred curves, from 

 which we have given what we think fairly representative 

 examples. Moreover, the curves analysed were in many 

 instances chosen from numerous examples, so as to represent not 

 one experiment, but several. We will now state the results 

 arrived at in somewhat more general terms. -.■-•i 



1 . At those pitches commonly employed in conversation the 

 sound consists essentially of the first two partials of the note 

 employed. The proportions between these partials may vary 

 between l : i'5 and i :o*5 ; perhaps even more widely. When 

 curves obtained by speaking were compared with those obtained 

 by singing the proportions were much more nearly alike than 

 could have been expected. The spoken o differed chiefly from 

 the singing J by a continuous change of pitch, and also by 

 running into an ii towards the end of the utterance. 



2. When is sung from g downwards the third, fourth, and 

 fifth partials appear in succession in such a way as to keep that 

 which is the highest prominent partial not far from b"^, which is 

 called by Helmholtz the characteristic tone of 0. Above g the 

 second and strongly reinforced partial lies in a region varying from 

 three semitones below b'\f to six semitones above it. 



3. Until we reach the lowest notes G and A all the partials 

 between the prime and the highest prominent partial are more 

 or less reinforced. Thus successively consists of the first 

 two partials, the first three partials, and the first four partials. 

 On very low notes we have insufficient data for very positive 

 conclusions, but the second partial seems to sink into insig- 

 nificance as the fifth comes in. 



4. The appearance of new partials as we descend the scale is 

 in some cases and with some voices gradual ; in other cases it is 

 singularly abrupt. 



Wave-form of Siins h' the same Vo'ce at Various Pitch's. 



Fig. I. 



Thisjis shown by Tables III. and IV., vhi:h give tha actual 



