723 
of Edinburgh, Session 1877-78. 
method suggested by Professor Tait was adopted. A charge was 
put upon the inducing ball of the pair on the stand, and the 
deflection read ; the charge was then divided by bringing an equal 
ball into contact with the inducing ball, the deflection read, and so 
on. The deflections were so nearly halved each time, that we may 
infer that they are strictly proportional to the charge on the inducing 
ball. We had arranged to verify our former observations in this 
matter last Thursday forenoon (27th June) ; but as the deflection 
on the scale always fell in the negative direction, and went to a 
great distance beyond the proper zero when the dividing ball was 
brought into contact, we gave it up. This effect was, doubtless, due 
to a strong negative electrification of the air ; for the thunderstorm 
came on immediately. 
2. On the Wave Forms of the Vowel Sounds produced by the 
Apparatus exhibited by Professor Crum Brown. By 
Professor Fleeming Jenkin, F.Pt.S., and J. A. Ewing, B.Sc. 
At a recent meeting of the Society, Dr Crum Brown exhibited a 
gutta percha bottle of irregular form, which, when applied as a 
resonance cavity to reeds of various pitches, gave very good imita- 
tions of certain vowel sounds. By closing certain apertures in the 
side of the bottle it could be made to say A (“father”), A (“ awe”), 
0 (“oh”), and I (“machme”). When the cavity was kept con- 
stant, and the pitch of the reed was altered, the same vowel con- 
tinued to be given. Dr Crum Brown was good enough to lend us 
the apparatus, in order that we might investigate the sounds given 
by the bottle in the same way as we have been investigating certain 
human vowel sounds, by obtaining and magnifying phonographic 
traces, and then subjecting them to harmonic analysis as far as the 
sixth partial tone. Of the vowels which the bottle speaks, 0 is the 
only one which we have fully examined in this way when spoken 
by the human voice, and we have confined our attention to it among 
the artificial vowels also. 
By using reeds of various pitches we have obtained curves or traces 
of the artificial O’s sufficiently good for harmonic analysis on the 
following pitches : — e, /jj, g, b , c', e'b, and e'. The pitch was in 
each case determined by measuring the length of the traces. The 
