352 Profs. Liveing and Dewar [Feb. 20, 



Fig. 5. 



From this experience I think we may consider that at an exhaustion 

 of a millionth of an atmosphere, air is an absolute non-conductor of 

 statical electricity. It is, therefore, legitimate to conclude that the 

 vacuum of interstellar space offers equal obstruction to the discharge 

 of electrified bodies, without necessarily interfering with their mutual 

 repulsion if similarly electrified. It is possible that in these facts an 

 explanation may be found of some obscure celestial phenomena. 



II. " On the Reversal of tbe Lines of Metallic Vapours." No. IV. 

 By G. D. Liveing, M.A., Professor of Chemistry, and J. 

 Dewar, M.A.. F.R.S., Jacksonian Professor, University of 

 Cambridge. Received February 12, 1879. 



In the experiments described in the following communication, instead 

 of introducing the substances to be observed in the metallic form into 

 our tubes, we have endeavoured to overcome, to some extent, the diffi- 

 culty of the presence of impurities by making use of reactions which 

 should generate the metallic vapours within the tubes. For this purpose 

 we liave generally employed the great reducing power of carbon and 

 of aluminium at high temperatures. 



In a former communication (" Proc. Roy. Soc," vol. xxvii) we 

 described the reversal of the two blue lines of caesiurn and the two 

 violet lines of rubidium by the vapours of those metals, produced by 

 heating their chlorides with sodium in glass tubes. It might be 

 doubtful from these experiments whether the absorption were due to 

 the metals or to the chlorides. To decide this question, we first tried 

 ceesium chloride by itself, heated in a tube such as we used before. ~No 

 absorption lines could be seen, although a good deal of the chloride 

 had been vaporized and distilled to the cool part of the tube. The 

 experiments were next repeated, "both with rubidium and caesium 



