408 
MR. W. CROOKES ON THE VISCOSITY 
and perforates the g'lass. The perforation is sometimes so fine as to cause the greatest 
trouble in finding the leakage., which only shows itself by the inability of the pump to 
give a good vacuum. Left to itself the logarithmic decrement slowly rises, the repulsive 
force of the candle increases to its maximum, and then slowly diminishes to zero, the 
logarithmic decrement continuing to rise till ultimately the internal and external 
pressures are identical. With a fine perforation it takes several days to go through 
these phases, and they occur with such slowness and regularity as to afford oppor¬ 
tunities for valuable observations. 
On one occasion I obtained a much higher exhaustion than 0'02 M. I could :iot 
measure it, but from the repulsion by radiation and the low logarithmic decrement 
(0'00537) I should estimate it at about O'Ol M. The terminals of the vacuum tube 
and wires leading to them were well insulated, and the full power of a coil giving a 
20 -inch spark was put on to it, At first nothing was to be seen. Then a brilliant 
green light flashed through the tube, getting more and more frequent. Suddenly a 
spark passed from a wire to the glass tube, and broke it, terminating the experiment. 
Since these experiments I have frequently got vacua as high, and even higher, but 
I have never seen one that would long resist the 20-inch spark from my large coil. 
VISCOSITY OF OXYGEN. 
666. The series of experiments with air show a complete history of its behaviour 
between very wide limits of pressure. It became interesting to see how the two 
components of air, oxygen and nitrogen, would behave under similar circumstances. 
Experiments-were therefore instituted exactly as in the case of dry air, but with the 
apparatus filled with pure oxygen. 
For these experiments the oxygen gas is prepared by the electrolytic decomposition 
of water containing sulphate of copper dissolved in it, a copper plate being used for 
