1885.] On Radiant Matter Spectroscopy. 415 



Elimination of Mercury Vapour from Vacuum Tubes. 



It is much more difficult than is generally supposed to keep mercury 

 vapour from diffusing into the experimental tubes. 



The following plan answers perfectly so far as my experiments have 

 yet gone : — Sulphur is first prepared by keeping it fused at a high 

 temperature till bubbles cease to come off, so as to get rid of water 

 and hydrogen compounds. It is then allowed to cool, and is pounded 

 and sifted so as to get it in the form of granules averaging a 

 millimetre in diameter. A glass tube, a centimetre in diameter and 

 about 2 feet long, is lightly packed for half its length with this 

 sulphur, and next about 2 inches of iodide of sulphur (I 2 S 2 ) is added, 

 and the rest of the tube is then filled up with sulphur. Ignited 

 asbestos is packed in at each end to keep the sulphur from blowing* 

 out whilst the vacuum is being made, or from being sucked through 

 when air is suddenly let in. This contrivance entirely keeps mercury 

 vapour from passing through, since the iodide of sulphur holds its 

 iodine very loosely, and fixes the mercury in the form of non- volatile 

 red iodide. A glass tube containing finely divided copper must 

 follow in order to keep sulphur out. With this blockade interposed 

 between the pump and experimental tubes I have been unable to 

 detect mercury vapour in any of the tubes, whether in the cold or on 

 heating them. 



The " Orange Band " Spectrum. 

 Since the date of my last paper I have devoted myself to the task 

 of solving the problem presented by the double orange band first 

 observed in 1881. With the yttrium experience as a guide it might 

 be thought that this would not be a difficult task, but in truth it 

 helped me little beyond increasing my confidence that the new, like 

 the old, spectrum was characteristic of an element. The extreme 

 sensitiveness of the test is a drawback rather than a help. To the 

 inexperienced eye one part of "orange band " substance in ten thousand 

 gives as good an indication as one part in ten, and by far the greater 

 part of the chemical work undertaken in the hunt for the spectrum - 

 forming element, has been performed upon material which later 

 knowledge shows does not contain sufficient to respond to any known 

 chemical test. 



Chemistry, except in few instances, as water-analysis and the 

 detection of poisons, where necessity has stimulated minute research, 

 takes little account of "traces;" and when an analysis adds up to 

 99*99, the odd 0*01 per cent, is conveniently put down to " impuri- 

 ties," " loss," or " errors of analysis." When, however, the 99*99 per 

 cent, constitutes the impurity and this exiguous 0*01 is the precious 

 material to be extracted, and when, moreover, its chemistry is 



