388 Dr. A. Schuster on the Spectra of Metalloids. [May 16, 



information to add on that point, and the wave-length of the lines will 

 be found in my letter. 



The Spectrum of the Negative Pole. — This spectrum has first been 

 correctly described by Wiillner. It consists of five bands, one of 

 which is too weak to be measured. The appearance of each band has 

 already been described (page 384). Careful measurements of the bands 

 have been taken. With regard to the explanation of the separate 

 spectra found at the negative pole in nearly all gases, I incline to the 

 view that they are due to separate molecular compounds which are 

 formed at the pole, The following experiments seem to support that 

 view. When the pressure is very small the spectrum of the negative 

 pole extends throughout that half of the tube which encloses the 

 negative pole, and which I shall call the negative half. If the current 

 be suddenly reversed the spectrum of the negative pole will still be 

 seen at first, in that part which was the negative half and now is the 

 positive half of the tube ; but it will gradually disappear and a perma- 

 nent state will be established, in which the spectrum of the negative 

 pole is, as before, only seen in the negative half. That it is the 

 reversal and not the interruption of the current which produces the 

 result is easily proved by interrupting the current and at once closing 

 it again the same way, when no difference will be seen. If, however, 

 the current be left interrupted for some time, say one minute, so that 

 any compounds which may have been formed in the negative half 

 may diffuse into the other half, and if then the current is closed either 

 the same or the opposite way, the negative spectrum will be seen at 

 first throughout the tube, but gradually disappear in the positive 

 half. 



If the current be rapidly reversed in succession, after a little while, 

 when the effect of the first reversal has disappeared, the permanent 

 state will always be established at once, and the spectrum of the 

 negative pole will appear only in the negative half. 



If after the last experiment the current be interrupted for some 

 time and then closed, the spectrum of the negative pole will at first be 

 seen throughout the tube, and gradually disappear in the positive 

 half. 



It is not quite easy to see the explanation of the last two experi- 

 ments. 



The experiments were all made in the Cavendish Laboratory, Cam- 

 bridge, and I am much obliged to Professor Clerk Maxwell, for the 

 kindness with which he has placed the resources of the Laboratory at 

 my disposal. 



