368 Lord Rayleigh on Experiments 



purity of a water surface. In a very good light a special 

 dusting might probably be dispensed with, the motion of the 

 surface being evidenced by inevitable motes. 



If the dust be applied in the first instance to a small central 

 patch, which is then touched internally with a very small 

 quantity of oil, the expansion of the dust in the form of a ring 

 is followed by a slight but unmistakable rebound. The effect 

 appears to take place when the surface is very clean to begin 

 with, and is then somewhat difficult of explanation. I am 

 disposed to think that it must be attributed in all cases to 

 initial contamination. This is concentrated in front of the 

 rapidly advancing ring, and has not time to diffuse itself 

 equally over the whole external area. Under the influence 

 of inertia the expansion of the central area may then proceed 

 so far that its tension becomes greater than that of the parte 

 immediately surrounding. 



Influence of Heat. 



For a lecture experiment the effect of heat is best 

 shown by holding a hot body near the surface of water con- 

 tained in a shallow vessel with a glass bottom. The hot 

 body may be the end of a glass rod heated by a flame, or 

 more conveniently a small spiral of platinum wire, rendered 

 incandescent at will by an electric current. The immediate 

 effect of the heat is to lower the tension of the part of the 

 surface affected ; but the visible result depends entirely upon 

 whether the surface be clean or otherwise. In the former 

 case the heated surface expands, and an outward current is 

 generated. This is rendered evident by the clearing away of 

 dust. But if the original contamination exceed a very small 

 quantity, a moderate expansion of the heated area brings the 

 tension again up to equality with that of the surrounding 

 surface, and there is no further action. In this case there is 

 no visible clearing away of dust under the hot body. 



Under favourable circumstances a very slight elevation of 

 temperature suffices. On July 28 a shallow tin vessel 8x5 

 inches, the lid of a biscuit-box, was levelled and filled with 

 tap-water from a rubber-hose, after a thorough preliminary 

 rinsing in situ. A little dust (sulphur) was then scattered 

 over, and the finger was brought underneath into contact 

 with the bottom of the dish. After about 20 seconds the dust 

 opened out, and a bare spot was formed over the finger of 

 about 1^ inch diameter. A spirit-flame, applied for a few 

 seconds under one end of the dish, cleared away the dust from 

 the larger part of the area. If when quiet was nearly restored, 

 a little fresh dust was applied, and the experiment with the 



