32 



Sir jSTorman Lockyer. On the Spark 



[Jan. 31, 



portion showing absorption on its more refrangible side, with fluffy 

 shading on its less refrangible side." 



Afterwards, when higher dispersions became available, the investiga- 

 tions of Messrs. Humphreys and Mohler on the effect of pressure on 

 spectrum lines* showed that the actual wave-length of a line was 

 increased by pressure ; thus Humphreys! states " the wave-lengths of 

 all fine and sharp lines, and also of the reversals of heart/ ones, increase 

 with increase of pressure around the arc, no matter how the lines 

 may spread out, symmetrically or chiefly towards either side." 



In the case of pressures of twelve atmospheres, a shift of scarcely 

 0"05 tenth metre was observed by Messrs. Humphreys and Mohler. 

 Eder and ValentaJ in their work on the spark spectra of argon and 

 sulphur under pressure obtained a displacement amounting to as 

 much as one tenth-metre. With flame spectra of the easily volatile 

 metallic salts, small displacements, averaging - 4 tenth metre, were 

 observed by Ebert,§ and were explained by him as being due to an 

 unsymmetrical broadening of the lines towards the red. 



Dr. Wilsing thought that such investigations suggested || " the direc- 

 tion which must be taken in the experiments for producing shifts of 

 lines without motion in the sight line, and ultimately for producing 

 double spectra." 



Wishing to avoid the experimental difficulties necessarily connected 

 with the employment of high pressures, he made use of the fact that 

 very high tensions are produced when electric sparks are discharged in 

 liquids. 



He employed a large induction coil, with a spark gap inserted in the 

 secondary circuit, in connection with a battery. With the passage of 

 each spark " a blinding discharge took place between the electrodes in 

 the water, giving a very intense continuous spectrum crossed by faint 

 lines." The discharge spectra in water and air were photographed on 

 the same plate with a spectrograph, the scale of the spectrum beings 

 about 50 mm. between A 4800 and A 4600, and the accuracy of the 

 determination of the wave-lengths of the sharp lines could be obtained 

 within a few hundredths of a tenth-metre. Further, several plates 

 were employed which were secured with a grating spectrograph of high 

 dispersion, and with a large prism spectrograph. 



Dr. Wilsing investigated in this way the spectra of the metals iron, 

 nickel, platinum, copper, tin, zinc, cadmium, lead, and silver, and 

 arrived at the conclusion that " there now occur displacements of lines 



* ' Astrophys. Joura.,' vol. 3, pp. 114-135, 1896; vol. 4, pp. 175-181, 249-262 

 (1896) ; vol. 6, pp. 169-232, 1897. 

 f 'Astrophys. Joura.,' vol. 6, p. 183. 



X ' Denkschriften der K. akad. der Wiss. zu Wien., 5 vol. 64, pp. 1-39, vol. 67, 

 pp. 97-151. 



§ ' Wied. Ann.,' vol. 34, pp. 34-90, 1888. 

 || 4 Astrophys. Jo urn.,' vol. 10, p. 115. 



