[CARR] AN EXTENSION OF PASCHEN’S LAW 163 
the existence of the relation which Paschen found to hold at high 
pressures. 
Owing to the special precautions taken by Peace to obtain accu- 
rate values for the spark potentials, it is possible to arrive at but one 
of two conclusions regarding the departure from Paschen’s law indi- 
cated by Peace’s numbers. Judging by the results, either the law 
ceases to hold when the critical pressure is passed or else the appa- 
ratus used by him in his experiments did not admit of an accurate 
measurement of the actual spark lengths corresponding to different 
spark potentials. 
A short discussion of the apparatus will reveal one considerable 
defect. ‘The object of the investigations of both Paschen and Peace 
was to determine the electromotive intensity requisite to cause dis- 
charge in a gas. Throughout the range of pressures investigated by 
Paschen the discharge always took place along the shortest distance 
ketween the spherical electrodes, and the electromotive intensity 
requisite to break down the gas was, therefore, directly proportional 
to the spark potentials obtained by him. At points below the crit- 
ical pressure, as Peace’s results indicate, discharge occurs more easily 
over a longer distance than over a shorter one, and if the values of 
the electromotive intensities necessary to break down a gas at differ- 
ent pressures are to be compared, it is necessary to know in each case 
not only the potential difference applied to the electrodes, but also 
the path between the electrodes along which the initial discharge 
occurs. 
To insure passage of the discharge over the same length of path 
Peace used plane parallel plates of very large diameter as electrodes; 
but while in this way he obtained a uniform field of considerable 
extent, and so was able to obtain an accurate measure of the electro- 
motive intensity between the electrodes, he failed to make certain 
that the path along which the gas initially broke down was always 
confined to the uniform part of the field. As mentioned in his paper, 
there was considerable tendency at low pressures to a brush discharge 
from the edges of the plates and this indicated a defect in his appa- 
ratus which apparently he did not completely eliminate. 
In the present paper an account is given of an investigation on 
the potentials necessary to produce discharge in a gas, with a form of 
apparatus which insured the passage of the discharge in a uniform 
electric field. 
With this apparatus the discharge potentials have been deter- 
mined for different distances between the electrodes over a range 
extending considerably above and below the critical pressure. The 
results of the investigation not only confirm the truth of the law 
