NovemBEr 8, 1918] 
particles of solid matter on the surface of the 
mercury were carried away by the evaporating 
mercury and moved about in the vessel in a 
very chaotic manner similar to the move- 
ment of molecules as postulated in the kinetic 
theory of gases and vapors. The stream of 
vapor which carried with it the particles was 
condensed upon the walls of the vessel and 
dropped back into the pool in the bottom of 
the tube. 
A simple tube for illustrating the phe- 
nomenon is shown in the figure. The tube is 
about ten inches long and one inch in diameter 
and contains a small pool of mercury. A small 
quantity of finely crushed material, such as 
colored glass or carbon is placed upon the 
surface of the mercury to form a layer two or 
three millimeters deep. Blue glass crushed in 
a mortar to give pieces about one half to one 
millimeter in size is found to be very satis- 
factory since particles of this size are easily 
visible. Particles of granular carbon or any 
other light material will do equally well pro- 
vided that it does not amalgamate with the 
mercury. The tube is then evacuated and 
sealed from the pump in the usual manner. 
The degre&pf vacuum is not essential provided 
that it s to prevent oxidation of the 
_ SCIENCE 
475 
mercury when the latter is heated, and thus 
to prevent the mereury from becoming sticky 
and adhering to the particles. The tubes ex- 
perimented with were evacuated to a pressure 
of a few thousandths of a millimeter. 
A gradual increase in the temperature of 
the mercury brought about by holding it over 
a Bunsen flame shows the following interest- 
ing phenomena: At a low temperature the 
particles begin to move about on the mercury 
surface. This movement of the particles gives 
the surface formed by them the appearance of 
a liquid agitated by convection currents. In 
this condition the particles have left the mer- 
cury surface and are moving about among 
each other close to the mercury in a layer 
having a rather well defined surface. A fur- 
ther slight increase in the temperature of the 
mereury causes some of the particles to leave 
the layer and to move about chaotically in the 
space above the surface similar to molecules 
which have left an evaporating liquid. At the 
same time the surface formed by the particles 
becomes indistinct and there is a gradual 
gradation of density of particles upward from 
the region just above the mercury. As the 
temperature continues to rise more of the 
particles leave the surface and also those 
which are moving in the space move to a 
greater height. Finally all of the particles 
leave the surface of the mercury and move 
about in the space colliding with each other 
and with the sides of the tube like the mole- 
cules of a gas. 
The phenomena just described are more 
easily produced by first heating the mercury to 
the temperature at which all of the particles 
have left the surface and then observing them 
while the tube gradually cools; the above-de- 
scribed processes then occurring in the reverse 
order and simulating a condensing vapor. Be- 
cause of the vacuum in the tube the tempera- 
ture which is necessary to cause all of the 
particles to leave the surface is not sufficient 
to make the top of the tube too warm to hold 
in the hand. In order to make the particles 
more easily visible it is desirable to provide a 
white background by painting or frosting one 
side of the tube. 
