1310 
shifts, those of BrLHaM') as to the spark-spectrum and those of 
Durrierp®) as to the arc-spectrum. 
In his above-mentioned paper the latter author has made a com- 
parison between his observations and those of Brrmam and this 
suggested the question to me which of the series of observations 
would agree best with the observed Zrrmar-effeet and whether per- 
haps in this spectrum a closer connection might be found. 
From the observation-material it is evident that there does not 
exist a simple proportionality between the two effects. 
We may however take together large groups of lines and see, 
whether we can remark something about these groups. As most of 
the lines, which can be taken into consideration for the investigation 
give triplets in the magnetic field, while only a few quadruplets 
and sextuplets are among them, a division into groups according 
to the magnetic decomposition is not well possible. 
The different behaviour of the lines at an inerease of pressure 
however suggests a division into groups. Corresponding to GaLE and 
ApDAMs*) we shall distinguish: 
Group I. Lines which are symmetrically reversed. 
Group IH. Lines which are asymmetrically reversed. 
Group [IL Lines which under pressure remain bright and narrow. 
Group IV. Lines which under pressure remain bright and sym- 
metrical, but which become wide and diffuse. 
Group V. Lines which under pressure remain bright, but which 
are widened asvmmetrically towards the red side of 
the spectrum. 
In this division of groups all lines will be taken up of which 
BrLHam has determined the group to which they belong and the 
intensity (1) for the spark-spectrum; then also the difference in 
wavelength between the two magnetic components which vibrate 
perpendicularly to the lines of force, measured in 0,001 A. UZ. 
The following 4 columns contain the difference in wavelength ex- 
pressed in 0,001 AU. per atmosphere increase of pressure. The 
first two (B, and B,) have been calculated from the observations of 
Birnam for 5 resp. 10 atm. pressure-increase and the two next ones 
(D, and D,) from those of DurrieLp for an increase of 10 resp. 
20 to 100 atm. 
1) B. G. Buuam. The Spark Spectra of Nickel under Moderate Pressures. Phil. 
Trans 214 A. p. 359. 1914. 
2) W. G. Durrrerp A Comparison of the Are and Spark Spectra of Nickel Pro- 
duced under Pressure. Phil. Mag (6) 30. p. 385. 1915. 
5) H. G. Gare and W. S. Avams. An Investigation of the Spectra of Iron and 
Titanium under Moderate Pressures. Astroph. J. 35 p. 10. 1912. 
