248 MR JOHN AITKEN ON THE 



the same time a clear space was formed on the hot surface, and rose up from 

 it, producing a dark plane in the ascending current (fig. 7). This clear space 

 was evidently entirely due to the heat in some way driving the particles away 

 from the hot surface. 



When working with this flat test surface it is necessary to be careful about 

 the adjustment of it in a vertical plane. If the surface leans either to the one side 

 or to the other, a clear space is, of course, formed on the side to which it inclines 

 by the separating action of gravitation, and gravitation also acts on the particles 

 on the other side, and tends to counteract the effect of the heat. Further, if 

 the surface is inclined enough, the gravitation effect overcomes the heat effect, 

 and destroys the dark space by causing the particles to fall towards the hot 

 surface. At the same time, the gravitation dark space on the under side 

 becomes thicker and thicker the more the plane of the test surface approaches 

 the horizontal. This instrument may be made capable of measuring the relative 

 effects of different temperatures, &c, by providing it with a scale to indicate its 

 angle with the vertical. The greater the angle at which the dark space is 

 visible the greater will be the repelling force. 



By the construction of the instrument, when placed vertically, the gravita- 

 tion effect is entirely removed. The dust particles can be seen coming straight 

 up, and no purified current coming from the under side (compare figs. 5 and 7). 

 The clear space begins to show itself with a very slight rise of temperature. 

 Indeed, it would appear that it is formed by the slightest rise of temperature, 

 as it always begins to be visible just when the temperature is high enough to 

 cause an ascending current, With a slight difference of temperature it is 

 extremely thin, and requires careful observation to detect it, but as the tem- 

 perature rises it becomes thicker and thicker. For the present I shall not enter 

 into the question as to why the dust particles move away from a hot body, but 

 shall leave the consideration of this subject till after describing some experi- 

 ments which seem to throw some light on the mechanism of these movements. 

 For the present I shall simply speak of it as repulsion due to heat. 



The following experiments will help us to understand the action of this 

 repulsion. Fix a piece of glass in front of, and parallel to, our flat test 

 surface, and at a distance from it of two or three millimetres. Glass is used 

 because it is transparent, and allows the illuminating beam to penetrate and 

 show us what is taking place at the different surfaces. If we now warm the 

 test surface, the dust particles all move away from it towards the glass plate, and 

 many of them attach themselves to the glass. ^ After a short time the glass 

 gets warmed by radiation, &c, from the hot test surface. If we now cool the 

 test surface a change takes place, the dust particles move away from the glass, 

 and crowd up towards the colder test surface. 



A better form of the experiment is shown at figs. 8 and 9. A glass plate 





