508 Prof. Tyndall on the Action of Free Molecules on 



rays which act upon the coloured vapours to pass freely 

 through it. A layer of dissolved iodine, on the other hand, 

 deprives the beam of its power of evoking sounds from either 

 iodine or bromine vapour. 



The rotation which produces the maximum effect is soon 

 ascertained by experiment. The sound is loudest when the 

 pulses succeed each other in periods which invoke the reson- 

 ance of the flask. I possess a hollow cone with a well-polished 

 rock-salt base, by which this point is well illustrated. Filling 

 this cone with chloride of methyl, while the base is turned 

 towards the source of heat, the apex of the cone being con- 

 nected by a tube with the ear, sounds of extraordinary in- 

 tensity are produced. Abandoning the ear-tube, the sound 

 can be heard at a distance. But to obtain this effect the speed 

 of rotation must be definite and constant. The maximum 

 sound once obtained, either the lowering or the heightening 

 of the speed rapidly enfeebles it. It is difficult by hand- 

 turning to keep the rate of rotation constant. Hence the 

 desirability of a mechanical arrangement which would ensure 

 the proper rapidity and necessary uniformity. One or two 

 motive powers have been tried, including a small steam-engine 

 and an electro-magnetic engine; but the arrangement has not 

 yet been brought to perfection. 



§ 9. Manometrk measurements. 



Some time before the visit of Prof. Graham Bell inXovember, 

 1880, I had inserted into my old experimental brass cylinder 

 a narrow tube of glass, which, being bent at a right angle a 

 few inches above the cylinder, could hold an index of coloured 

 liquid in its horizontal portion. I had long known that the 

 absorption of radiant heat must be accompanied bv the ex- 

 pansion of the absorbing body, but thought that such expan- 

 sion would furnish only a rough measure of the absorption. 

 With ordinary sources of heat I found the expansion small, 

 even when sulphuric ether occupied the experimental tube ; 

 but when a pair of stout carbons, rendered incandescent by a 

 Siemens machine, were employed as a source, the liquid index 

 was driven forcibly out of the narrow glass tube. 



The experimental tube, however, was but a rude mano- 

 meter; and I therefore sketched and described to my assistant 

 at the time, with a view to its construction, a handier instru- 

 ment. The apparatus was to consist of a short tube with 

 rock-salt ends, capable of being exhausted and filled with any 

 required gas or vapour. Through this tube it was proposed 

 to send a concentrated calorific beam, whose action on an 

 absorbent gas or vapour should be declared by the depression 

 of a liquid column in one leg of a U tube, and its elevation in 





