FORMATION OF SMALL CLEAR SPACES IN DUSTY AIR. 249 



A, 12 cm. long and about 4 cm. broad is attached by means of cement, near its 

 upper end, to a metal tube, to enable us to heat it while in the dust-box. 

 Another plate of glass B, of the same size as A, is placed opposite and parallel 

 to it, at a distance of about 5 mm. The plate attached to the tube is first put 

 in its place in the box, and after it has acquired the temperature of the air, 

 the other plate B is warmed and put in its place, opposite to A, as shown in 

 sketch. The box is now filled with dust. If we now carefully examine the 

 air between the two glass plates, we shall find that the warm plate B (fig. 8) is 

 bounded on each side by a clear space, its high temperature having driven all 

 the dust particles to a distance, while the other plate has no clear space round 

 it. Now let us put a little warm water into the tube to heat the upper part of 

 the cold glass plate A, and note the change in the distribution of the dust. 

 As before, the lower part of B is bounded by a clear space (see fig. 9), but the 

 upper part of A being now warmer than B, the dust is driven from A towards B, 

 and a clear space opened in front of the hot part of A, while the clear space 

 formerly in front of the upper part of B is closed. The heat has thus caused 

 the dust particles to move across the direction of motion of the air. These 

 experiments have been made with different dusts, and always with the same 

 result. 



For the purpose of studying the effects of higher temperatures than that of 

 boiling water, a fine platinum wire was fitted up inside the dust-box and heated 

 by means of a small bichromate battery. The arrangement of wire which I 

 prefer for this purpose is made by bending it into a U-shape and bringing the 

 two legs close together, say one or two millimetres apart. The wire is placed 

 horizontally in the dust-box, with the bend to the front, and the legs at the same 

 level, the two copper wires to which it is attached being carried backward and 

 out of the box. By this arrangement a clear end-view is obtained all round 

 the wire, and the effect of the heat conveniently observed, and further, the 

 Avire doubled in this manner, tells us more than a single wire can. 



In experimenting with this arrangement of apparatus, the results are as 

 varied as the dusts employed. Each dust gives a different size of dark plane 

 for the same temperature. The previous experiments with less intense heat 

 seemed to point to repulsion as the cause of the clearing away of the particles. 

 If this were the case, it seemed very unlikely that some dusts would be 

 repelled further away than others, at least to the extent that actually took 

 place. To see if repulsion was the explanation in this case also, instead of a 

 single wire, which I used in my first experiments, I doubled the wire into a 

 U-shape, as already explained, and placed the length horizontally, with the legs 

 at the same level. When this wire was heated in the sal-ammoniac or in sulphate 

 dust, it was at once evident that repulsion was not the cause of the dark plane 

 in these dusts. With either of them, when the temperature of the wire was not 



