2U 



Mr. G. P. Thomson on the 



that some of the emission may be due to the partial separation 

 of an electron from the atom and its return to the normal 

 position. 



Molecules. 



It was never found possible to get a state of discharge in 

 which the positive rays were exclusively molecular. At 

 very low pressures indeed, the atomic parabola became 

 almost invisible in some cases, but at these pressures the 

 light in the observation-chamber was too faint for spectro- 

 scopic examination. The best that could be done was with 

 the larger discharge tube (fig. 3). With the dark space 

 about half across the tube this showed a marked excess of 

 molecules over atoms as determined by the brightness on 

 the willemite screen. Since it is known that such a screen 

 is considerably more sensitive to atoms of hydrogen than to 

 molecules, it is probable that in these experiments there was 

 really a very large excess of molecules. A spectrograph 



Ffc. 3. 



Approx. limit 

 of dark space 



Insulated 



taken under these conditions is shown in PI. IV. No. 2, the 

 dark space being as shown in fig. 3. It will be seen that 

 the second spectrum is clearly shown, though the H^ line is 

 still the brightest on the plate. This is not surprising when 

 we consider that the light of the Balmer series is con- 

 centrated into very few lines relative to that of the secondary 

 spectrum. The brightness of the second spectrum in this 

 case, as compared with the former plate — in which indeed 

 it was invisible, — leaves no doubt that the secondary spectrum 

 is caused by the hydrogen molecule. This result is con- 

 firmed by the results of experiments with another discharge 

 tube (fig. l),in which increasing proportions of molecules 



