1670 
V. _ Preliminary measurements on liquid miatures 
of oxygen and nitrogen. 
1. The viscosity of liquid air. 
It was resolved to begin by a determination of the viscosity of 
liquid air, in order to test the working of our apparatus in a tech- 
nically simple case. 
The liquid, freshly prepared in the laboratory, was introduced 
into the vacuumglass up to the desired level (about the mark on 
the steel pin), the vacuumglass being surrounded by a second one, 
which was also filled with liquid air up to a somewhat higher 
level. Owing to the radiation from the cap of the apparatus *) some 
evaporation took place at the surface of the liquid, by which the 
proportion of nitrogen diminished and the boiling point rose a little 
in the course of a day; during one experiment the change was 
however, imperceptible. A more serious inconvenience arose from 
the convection-currents which accompanied the evaporation giving 
irregularities in the vibrations of the sphere: the disturbance was 
avoided by stirring the liquid immediately before an experiment: 
this was effected by operating a hand suction-pump attached to the 
tube which carries off the vapour, by which means bubbles were made 
to rise through the liquid from a syphon-tube which reached down 
into the bottom of the liquid.*) This operation produced a small 
fall of temperature (a few hundredths of a degree), which recovered 
itself during the experiment, as shown by a small diminution of 
the logarithmic decrement during the experiment. The mean of the 
temperatures before and after the experiment was taken as the 
temperature corresponding with the mean decrement. 
The temperature was measured by means of an oxygen vapour- 
pressure-thermometer. A small vessel of about 1 ec. of the shape of 
a small tube 7 ems. long was placed at the level of the swinging 
sphere, between the spherical envelope and the wall of the vacuum- 
vessel (see figure) and by means of a glass capillary, a german silver 
tube soldered in the cap and a steel capillary was connected to a 
manometer 7’ which is of much the same design as that described 
by Braak *). The thermometer was filled with pure oxygen, the vapour- 
1) This radiation we hope to be able in future to obviate almost completely 
by the introduction of screens inside the vessel. 
2) This tube which can be seen in the figure is used in emptying the apparatus 
after the experiment; during the experiment it was connected with a small open 
gauge (Ma in the figure) which indicates any excess of pressure of the vapour in 
the apparatus above atmospheric pressure. 
3) Comm. 107a. The modifications introduced in its construction since then will 
be described on a different occasion. 
