204 THE ROYAL SOCIETY OF CANADA 



corroborate the result obtained we made five new photographs of 

 the mercury spectrum by means of a glass prism spectrograph. 

 These are shown in Fig. 20. The source of the light used was a 

 quartz mercury arc lamp which was operated with a current of about 

 4 amperes, under a potential difference of about 40 volts. The slit 

 used was 0.2 mm. wide. The plates were stained with Dicyanin A. 



Photograph (a) shows the mercury spectrum as obtained with 

 an exposure of 1 minute on a panchromatic plate. The photograph 

 has the usual appearance and extends into the red up to X 6908A. 

 The next (b) which was made on a stained plate with an exposure of 

 20 hours, is not so rich in lines as the first, but it shows already a 

 slight trace of the line X 10140A. 



Next (c) with exposure of 40 hours showing very clearly the line 

 X 10140A has one very interesting feature. While all the strong lines 

 appeared as black ones, two lines X 6717A and X 6908A are white. 

 The fourth (d), taken with 60 hours exposure, has the two above 

 mentioned lines turned black, and at the same time shows a line 

 X 7729A not previously registered. The fifth {e) photograph, although 

 of 80 hours exposure, is more like the third (c); apparently some 

 unknown cause reduced the intensity of the light and twice as long 

 an exposure as in the case of (c) only compensated for this reduction 

 of the intensity. Thus the last photographs clearly showed that the 

 mercury line X 10140A, as registered by Paschen^ using a linear 

 thermopile and galvanometer can be photographed and is one of the 

 strongest lines in the mercury spectrum. It may be stated that in 

 identifying the line X 10140A a calibration curve, a reproduction of 

 which is shown in Fig. 21, was made for the spectrometer. In making 

 this curve wave-lengths commencing at X4358A for mercury were 

 used as ordinates and displacements in mms. on the photographic 

 plates from the line X 4358 were taken as abscissae. 



An interesting point was brought out, namely that long exposures 

 give white lines on the negative and black ones on the positive. This 

 fact finds its explanation in the phenomenon noticed by Waterhouse 

 and applied by Millochau^ that infra-red rays destroy photographic 

 action, on a plate which has previously been exposed to faint light. 

 As the above photographs show it must be stated that visible rays 

 possess this property too. 



As to the plates used they had not been "solarized" but strong 

 visible lines acting on the plate produce this "solarization," and later 

 destroy it, appearing as white on the plate. Weak visible lines with 



^Paschen, Ann der Phys., Vol. 27, p. 559, 1908. 

 »Millochau, Comptes Rendus, 144, 725, 1907. 



