PR. W. GEOFFREY DUFFIELD ON THE 



atmospheres, which are not very concordant, it is apparent that from 40 to 200 

 atmospheres there is a general downward tendency, indicating, if the figures are 

 ivlinhl.-. that there is a more rapid rate of increase of the displacement with the 

 j.ivssure at low than at high pressures. Group III. shows this downward tendency 

 strongly, and so does, to a marked degree, the line jl in Group I., which line, it 

 should 1 mentioned, was measured both in Set A and Set B of the photographs with 

 concordant results. 



There is thus some reason for believing that there is a departure from a strict 

 linear relationship in the direction already indicated. 



A study of the negatives themselves reveals the difficulties to be encountered in 

 measuring the displacements and in fixing upon the most intense portions of the lines, 

 and it is at once admitted that the downward tendency of two or three of the lines 

 should not be allowed to disturb our belief in the simpler relation ; but the fact that 

 all the gold lines, with the exception of ll, present the same feature, taken in 

 conjunction with the fact that the copper lines are similarly affected (if their 

 behaviour up to 200 atmospheres be taken into account), provides strong evidence 

 that though the relation between the displacement and the pressure is approximately 

 linear, there is a slightly greater increase of the displacement with the pressure at 

 low than at high pressures. 



(7) Displacement and Reversal of Lines. The displacement diagrams for iron 

 suggested a departure from the linear relationship at those pressures at which a large 

 number of reversals appeared. No reversals have been observed in the spectrum of 

 the gold arc, so that the slight departure from a linear relationship, discussed in the 

 preceding section, cannot be associated with the phenomenon observed in the 

 iron arc. 



7. Changes in Relative Intensity under Pressure. This is not so prominent a 

 feature of the gold arc under pressure as it is of the copper and silver spectra ; in the 

 former, all the lines belonging (in the region investigated) to KAYSER and RUNGE'S 

 series vanished under pressure, and well-marked changes also took place in those 

 lines which were not members of any particular series, while in the silver arc the line 

 spectrum vanished completely under pressure. 



>fall the lines examined in the spectrum of gold, only one line, ol (X = 4241-99), 



, is obliterated at an early stage in the increase of pressure and is not 



25 atmospheres; the other lines usually preserve their relative 



th but slight modifications, which can, however, be generally associated 



the amount of broadening that the line undergoes. A line that remains 



1 narrow under pressure is likely to have a mean intensity greater than 



wing a large area. The intensity would be best expressed by 



the evaluation of ,' rfX, where ,' is the intensity at any wave-length X, and X, and X, 

 the extreme limits of the range of wave-lengths comprised in the broadened line. 



