582 
Proceedings of the Royal Society 
iron bar employed by Forbes, and reproduced, almost exactly, 
Forbes’ results for that metal — the most notable feature of which, 
besides the first really trustworthy determination of absolute con- 
ductivity, was the discovery that the thermal conductivity of iron 
is diminished by heating, and is very nearly proportional to the 
reciprocal of the absolute temperature. 
None of the other metals tried by the author, viz., copper (two 
specimens, selected as being specially good and specially bad con- 
ductors of electricity), lead, and German silver, gave at all analogous 
results. Their absolute conductivities have been determined ; and 
the changes produced by heat in these conductivities were, where 
any, very slight, and uniformly in the direction of improved con- 
ductivity at higher temperatures. 
A second part of the paper is promised, to contain comparisons 
of thermal and electric conductivity in the same substance, with an 
investigation of the “ Thomson effect.” 
2. On the Wave Forms of Articulate Sounds. By Professor 
Fleeming Jenkin and Mr J. A. Ewing. 
In this paper the authors gave a preliminary account of experiments 
made with the help of the phonograph exhibited at the last meeting. 
The following results have been obtained : — 
1. The vowel sounds can be produced by maintaining the relative 
pitch of the simple tones of which they are composed constant, 
although the absolute pitch of those simple tones may vary greatly. 
The authors found that whether they turned the barrel of the 
phonograph at a greater speed or at a less speed than that which 
had been maintained while the human voice produced the indenta- 
tions, the several vowels were given back w r ith equal distinctness. 
This experiment proves that within certain limits a given wave 
form in the tinfoil produces a given vowel independently of the 
rapidity with which it passes the pricker of the vibrating disc. 
It proves therefore that within these limits a given group of partials 
is competent to produce a given vowel effect, whatever be the 
absolute pitch of the prime tone, and of the relative partials. This 
result is of much importance, since Willis, Wheatstone, Donders, 
