FORMATION OF SMALL CLEAR SPACES IN DUSTY AIR. 247 



a clustless envelope. The two currents of clear air which started from the 

 under side of the tube, reunite at the top after passing round the sides, 

 and ascending in the centre of the upward current, form a well-marked dark 

 plane (see fig. 5). Here again gravitation seems to be the principal cause of 

 the distribution of the particles. This certainly is the case when the difference 

 or temperature is very slight, but we shall see that, as the temperature rises, the 

 gravitation effect bears a less and less proportion to the heat effects, which we 

 shall presently consider. It will be as well to note here the difference in the 

 clear space surrounding the tube in this case and when cold was applied, as 

 shown in figs. 2 and 5. When cold was applied (fig. 2), the dark space was only 

 on the under side of the tube ; but with heat it is all round the tube, because it 

 has its origin in the air under the tube. 



When making these experiments a somewhat peculiar effect was often 

 noticed, which seems worth recording, as it forms a good illustration of the 

 influences at work here. If, after the tube had been warmed and a well- 

 marked dark plane formed over it, no more hot water was added, and 

 the tube allowed to cool, the upward current became sluggish after a time, 

 and the dark space presented the appearance shown in fig. 6. The two sides of 

 the tube now differed. The left side was bounded by a clear space, which 

 ascended as before, but on the other side, the dark space did not continue to 

 the top of the tube. As shown, the particles here came into the dark space 

 and obliterated it. The explanation of this peculiar effect, which so often 

 repeated itself, is this. The falling temperature had allowed the current on the 

 right side to become so slow that gravitation had time to act on the particles 

 after the current turned to the upper side of the tube, and the particles had 

 time to fall through the clear space before they were carried into the ascending 

 current over the tube. In other words, gravitation undid on the upper part over 

 the tube what it did at the under. The left side of the tube continued to keep 

 its clear space, because the light used for illuminating it was focused on this 

 side; it, therefore, was slightly warmer than the other. 



Gravitation, while it explains the formation of the dark plane in such cases 

 as above described, where the difference of temperature is slight, is evidently 

 not the whole explanation. Gravitation can obviously have little to do with 

 the formation of the dark plane formed over a thin wire, as the time occupied 

 in horizontal movement when going round so small a body is not enough for it 

 to have any appreciable effect. In order to study the effects of heat apart from 

 those of gravitation, the tube with the flat surface fixed on it, employed in the 

 experiment with cold, was used, as it eliminates the gravitation effect and 

 shows the heat effect alone. Fixing this piece of apparatus in the smoke-box 

 with the test surface carefully adjusted in a vertical plane, heat was slowly 

 applied to it. An upward current at once started, and it was noticed that at 



VOL. XXXII. PART II. 2 T 



