34 



Mr. W. Crookes on Repulsion 



[Jan. 17, 



i.e. the black side is repelled. If one of the outer corners of each 

 plate is turned up at an angle of 45°, so as to keep the blacked surface 

 on the concave side, the positive rotation is either diminished, stopped, 

 or converted into negative rotation, according to the amount of surface 

 of the plate which has been turned up. 



When the plates are kept flat, but the supporting arms are bent so 

 as to present more of one side than of the other to the bulb, as shown 

 in fig. 1, the fly rotates under the influence of radiation in the direction 



Fig. 1. 



of the arrows, even when there is no difference between one surface 

 and the other. It is shown that the favourable presentation of the 

 surface of the vanes to the inside of the bulb has more influence on the 

 movement than has the colour of the surface. Experiments are 

 described with radiometers of the form shown in fig. 1, and made of 

 thick and thin mica, pith, aluminium bright on both, sides, and alumi- 

 nium blacked on one side. The action of light and of dark heat is 

 given. The negative rotation set up in the fly when it is cooling from 

 a high temperature, and the anomalous behaviour of the " favourably 

 presented " radiometers when immersed in hot air or hot water, are 

 examined. It is found that when a hot metal ring is applied to the 

 equator of the bulb, the direction of rotation is always positive ; and 

 that when a hot ring of a smaller diameter is applied to the top or 

 bottom of the bulb the direction of rotation is always negative. The 

 direction of movement when the hot rings are removed and the fly is 

 cooling is positive with the aluminium vanes, and negative with the 

 thin mica and pith, vanes. 



The positive rotation, when the bulb is heated equatorially, is inde- 

 pendent of the material of which the fly is made. It is caused by the 

 hot ring of glass generating molecular pressure, which radiates towards 

 the centre, and strikes the sloping vanes, driving them round as if a 

 wind were blowing on them. The other movements of these " favour- 

 ably presented " radiometers are explained in the paper, but it is 

 difficult to reproduce the explanations in abstract without the aid of 

 the diagrams which accompany the paper. 



Having investigated the simplest form of favourably presented 



