neighbourhood of strongly Illuminated Bodies. 221 



glass films about '0003 inch thick usually show no dark coat 

 at all, and they either set up no convection-currents, or those 

 which they set up are exceedingly sluggish. Strips of ordinary 

 glass give good coats, the one on the illuminated side appear- 

 ing first. It seems as if very thin films were incompetent 

 to absorb enough radiation. (It has been incidentally noticed 

 that these glass films, when freshly blown, adhere together 

 when placed in contact. After standing in the air for a few 

 hours this does not take place.) 



A lump of rock-salt in ordinary air exhibited dark coats 

 and a plane, but this might be due to moisture on the surface. 

 A good plate of clear rock-salt, in air thoroughly dried with 

 phosphoric anhydride, behaved just like the thin glass films, 

 giving no coats and doubtful convection-currents. A plate of 

 mica 0*002 inch thick absorbs sufficient radiation to exhibit 

 the effects, though when the sheet was horizontal the coat on 

 its upper surface was very badly marked. Mica showed an 

 unexpected result, which is described later on. 



Horizontal glass tubes filled with smoke, and immersed in 

 the beam, have been examined by looking along the axis. 

 The air inside is nearly stagnant, but the smoke collects in 

 the axis of the tube, leaving a clear dark space or ring all 

 round. The smoke settles also more slowly when the glass 

 is thus warm than when it is at the temperature of the 

 air. • 



(c) Nature of Surface. — Of three copper rods, one was 

 tightly covered with blotting-paper, one with cotton wool, and 

 the third left uncovered. They were all smoked with cam- 

 phor-black, and exposed to the beam : the cotton-wool surface 

 showed slightly the broadest plane. But the effect of such 

 variation of surface is not marked, and is probably insignifi- 

 cant. Other experiments made with smooth electric-light 

 carbon, and carbon roughened with sand-paper, led to the 

 same conclusion. 



2. Experiments made on Solids at different Temperatures. 



To determine roughly the heating-power of the beam of 

 light used, and to see at what temperature the dark plane be- 

 came distinctly visible, a thermometer with its cylindrical 

 bulb blackened with camphor-smoke was inserted horizontally 

 into the box (fig. 5), and used as the solid to be examined. 

 When the smoke was introduced,, the thermometer read 21°. 

 A few seconds after the light had been turned on, the dark 

 plane was visible, and the thermometer read 21 0, 2. At 21°*5 

 the plane was good, but there was no coat visible to the eye. 

 At 22° the coat appeared, and at 23° it w r as distinct. The 



