FORMATION OP SMALL CLEAR SPACES IN DUSTY AIE. 251 



vapours ; in the wider space b, both the chloride and sulphate dusts are 

 vaporised, and we have nothing visible save the indestructible dust ; in the 

 next space c the chloride is vaporised, and there are present the sulphate and 

 indestructible dusts ; while in the space d all the dusts are present, but dry, 

 the condensed water being evaporated. 



Conclusion. 



The conclusion we have arrived at from these experiments is, that for the 

 formation of the dark plane in dusty air, there are various causes which may 

 be classed under the following heads : — With cold, producing the downward 

 dark plane, we have — 1st, the distributing effect of gravitation; and 2nd, the dis- 

 appearance of the particles by evaporation, when falling into a space rendered 

 dry by condensation produced by cold. With heat, producing the upward dark 

 plane, we have — 1st, the distributing action of gravitation; 2nd, the distributing 

 action of repulsion due to heat ; 3rd, evaporation of the particles ; and 4th, dis- 

 integration of the dust. In the last two cases the dust is rendered invisible by 

 the heat changing it from the solid light-reflecting condition to the transparent 

 gaseous state. 



Effect of Centrifugal Force. 



We may here ask ourselves, Are these the only ways in which the dark 

 plane may be produced ? It is, of course, impossible to give a definite answer 

 to such a question. There are, no doubt, other ways in which it seems possible 

 that this phenomenon might be produced, and it seemed worth while to consider 

 Lord Bayleigh's suggestion as to the effect of centrifugal force. On consider- 

 ing the action of this force in the experiments described, it is evident that it 

 can have but little to do with the distribution of the particles, because the air, 

 in rising and passing round the wires and tubes, is curved first in one direction, 

 and before it again takes up its original direction of motion, it is curved to an 

 equal amount in the opposite way. So that whatever sifting action the centri- 

 fugal force may have at the one part of its course, will be undone at the other. 

 I, however, thought it worth while to arrange an experiment, to see if the par- 

 ticles really were thrown out by centrifugal force at any part of their passage. 

 With this object, I fixed inside the dust-box a piece of thin sheet metal, with 

 its plane vertical. Arrangements were made so that a current of dusty air was 

 caused to flow down the one side of the plate, round the lower edge, and up 

 the other side. In this way the air was caused to curve through, an angle of 

 180°, and no curving in the opposite direction took place. When this was 

 done, it was seen to be possible to give an appearance very like as if the cen- 

 trifugal force did throw the particles away from the centre of motion. In front, 

 md just above the lower edge of the plate, there was formed a clear space, 



