346 
PHYSICS: L. B. LOEB 
that for the negative ions such a break up did occur. It was conse- 
quently thought of importance to see if such a breaking up, or abnormal 
increase of mobility, of either ion could take place in high electric fields 
at ordinary pressures. It was considered particularly important to 
test this point for the case of the positive ion. If in this case an ab- 
normally high mobility were found, only one conclusion could be drawn 
as to its significance, namely, that the positive ion consists of a cluster. 
In the case of an abnormally high mobility of the negative ion, the 
presence of the free electron might make the result ambiguous. It was 
with this end in view that the work was undertaken at the suggestion 
of Prof. R. A. Millikan. 
Measurements were made by means of the Rutherford^ alternating 
current method. With this method the problem resolves itself into one 
of obtaining high field strengths, e.g., 5000 volts per cm., with alternat- 
ing fields of frequencies of about 5000 cycles per second. Such fre- 
quencies are demanded by the high fields in order to make the distance 
between the plate and the gauze (i.e., the distance that the ions tra- 
verse in one-half cycle), small enough to keep the field between the plate 
and gauze uniform. These high frequencies were obtained from the 
resonance circuit of a Chaffee^ arc. This arc gives nearly undamped 
sine waves, and is comparatively easy to operate. Using a tertiary 
circuit as seen in the diagram (fig. 1), an alternating potential difference 
of 4400 volts at a frequency of 7670 cycles per second was obtained 
from the arc. The frequency was determined by photographing the 
spark in a revolving mirror. The mobility-measuring apparatus is 
similar, as the diagram shows, to that used by other observers. The ions 
were generated by ionium on a disc / attached to the bottom of the 
gauze. The whole apparatus was placed inside a silvered bell jar 
which could be exhausted. Since neither the potential difference nor 
the frequency could be conveniently varied, the measurements were 
taken by varying the distance between the plate and the gauze. To do 
this, the apparatus was so arranged that the upper or electrometer 
plate could be raised or lowered from the outside of the jar. The 
distances between the plate and the gauze were plotted as abscissae, 
and the electrometer deflections as ordinates. The point where this 
curve intersected the axis of abscissae was taken as the critical distance 
which the ions could just traverse in the time of one-half period of 
oscillation. From the value of this distance J, the mobility U of the 
ion is obtained at once with the aid of the equation U = wNd^/E, in 
which is the frequency and E the maximum value of the potential. 
The measurements were made in air, which was carefully dried 
