464 
PROF. J. W. NICHOLSON AND DR. T. R. MERTON ON THE 
with the total width of the base. It is therefore necessary that the dispersion of the 
spectroscope should be sufficiently great in relation to the width of the lines. In the 
present investigation the lines measured have been of such a breadth that this 
condition is amply fulfilled, and, moreover, the photographs show that this is the case. 
At the maximum in the base of the wedge the image is spread out. If, however, the 
dispersion of the spectroscope is sufficiently great this spreading is no longer visible 
at the edges of the base of the wedge. An exaggerated drawing of the effect of 
irradiation on the wedges is shown in fig. 1. In “ A ” the effect of irradiation is seen 
at S, but at P and Q, which represent the boundary of the wedge and the 
“ recognisable density,” the effect is no longer visible. In “ B ” it is evident that the 
B 
effect of irradiation is not eliminated, and that a higher dispersion must be employed if 
reliable measurements are to be made for the line in question. 
(IV.) Experimental. 
For the production of the spectra investigated condensed discharges were employed 
from an induction coil capable of giving a ten-inch spark in air, and in parallel with 
a condenser having a capacity of 0'0025 microfarad. In the case of hydrogen the 
condensed discharges were passed between aluminium or platinum points in a glass 
vessel containing hydrogen at atmospheric pressure. In the case of helium a vacuum 
tube containing the gas at a pressure estimated at somewhat above a millimetre of 
mercury was employed, and in this case a spark gap was put into the circuit. The 
