RADIANT HEAT, AND ITS CONVERSION THEREBY INTO SOUND. 335 
so as to diffuse the vapour in the air above the liquid. The bulb is held, so that the 
point of maximum concentration of the beam falls upon it. 
With this apparatus I have tested more than once the sounding power of ten gases 
and of about eighty vapours. As a sound producer chloride of methyl is supreme. It 
is however closely followed by aldehyde, olefiant gas and sulphuric ether, the two latter 
being very nearly equal to each other. The volatility of the liquid from which the 
vapour is derived is of course an important factor in the result. For, however high the 
inherent capacity of the molecule as an absorber may be, if the molecules be scanty in 
number the effect is small. Feeble vapours, on the other hand, may to some extent 
atone by quantity for the inherent weakness of their molecules. A few examples will 
suffice to show how the specific action of the molecules over-rides the effect of volatility. 
Bisulphide of carbon with a boiling point of 43° C. is less powerful than acetic ether 
with a boiling point of 74°. Tetrachloride of carbon boils at 77° but its sound by no 
means equals that of acetal which boils at 104°. Chloroform with a boiling point of 
61° is less powerful as a sound producer than valeral with a boiling point of 100°, or 
even than valerianic ether with a boiling point of 144°. Cyanide of methyl boils 
at 82°, but produces less sound than acetate of propyl with a boiling point of 102°. 
In the experimental tube, these vapours follow, as absorbers, the order of their sounds. 
When tested in liquid layers they follow the same order. I have examined about a 
score of liquids with boiling points varying from 163° to 308°. At ordinary tempera¬ 
tures the vapours of these liquids were practically inaudible. The liquids being plunged 
in a bath of heated oil the vapours so produced emitted, for the most part, powerful 
sounds. The measured absorptions of a sufficient number of substances, in relation to 
their sounding power, shall be tabulated immediately. 
The fact is worth a passing reference that the chlorides of the elements appear one 
and all to be feeble sound producers, because they are one and all feeble absorbers of 
radiant heat. Many years ago I had found them highly diathermanous, and accepted 
chloride of sodium as representative of the class. Silicium chloride, for example, though 
very volatile is weak as a sound producer. Tetrachloride of carbon, and terchloride of 
phosphorus are also volatile, but not strong. Bichloride of tin, chloride of arsenic, 
chloride of titanium, and chloride of sulphur, are all feeble sound producers. In these 
three cases the boiling points are high, but non-volatility is not the cause of the weak¬ 
ness, for when the vapours are raised, by heating their liquids, almost to the pressure 
of the atmosphere, they are still but feebly sonorous. Whatever then the condition 
may be which renders these substances pervious to radiant heat, it appears to be a 
condition common to them all. 
In experiments with the chlorides care must be taken to abolish all fumes. 
Bichloride of tin sounds loudly with fumes, but is weak without them. Simply 
heating the upper part of the flask frequently suffices to lower the sounds from 
loudness almost to silence. 
