MR. E. G. BILHAM ON THE SPARK SPECTRUM OF 
3(54 
For most of the lines of wave-length shorter than X 3600 it is impossible to tell from 
examination of the eleven-atmosphere plates whether they belong to classes I. or II. 
The appearance of these lines under pressure is such that this region looks like a 
continuous spectrum crossed by absorption lines. On the six-atmosphere plates, how¬ 
ever, the emission lines are visible. 
Class V. displays the most striking features. The lines of this type become broad 
bands under pressure, with wings extending considerable distances towards the red. 
o 
A width of ten to twenty A. U. is not uncommon. A large proportion of the lines 
of wave-length greater than X 4400 behave in this way, so that a curious fluted 
appearance is imparted to the spectrum. On some plates the strongest lines of class 
V. are faintly reversed, so that they qualify for class II. The appearance is, however, 
radically different from the ordinary members of that class. 
In addition to the five well-defined classes described above, there is some indication 
of a sixth, analogous to the fifth, except that the winging extends towards the 
violet. The lines 4752'58 and 5035‘55 appear to behave in this way. 
5. Modifications in Intensity. 
The radical changes in the appearance of the lines, produced by increase of pressure, 
make it a difficult matter to draw any conclusions with regard to changes of intensity. 
In the case of the stronger lines there seems to be very little alteration. An examina¬ 
tion of the fainter lines, however, has yielded the following conclusions :—- 
(1) Enhanced lines are decreased in intensity. In some cases very little trace of 
the enhanced line is to be seen in the pressure-spectrum. This is particularly notice¬ 
able for a group of enhanced manganese impurity lines in the region X3400 to X 3560. 
The strong nickel enhanced lines 3576‘91 and 3769'62, however, are not greatly 
affected. 
(2) Lines which are relatively stronger in the arc than in the spark are increased 
in intensity. 
(3) Lines due to oxygen and nitrogen are completely eliminated. These results 
bear a striking analogy to the effects of including self-induction in an ordinary spark 
circuit. The general conclusion is that in the spark under pressure we get an 
approach to the arc condition. 
The following lines are also greatly reduced under pressure: 3668'35, 374475, 
443777, 443775, The line 452076 is notably strengthened. 
