806 



SCIENCE 



[N. S. Vol. XXVIII. No. 727 



Our preliminary experiments showed that it 

 was very difficult to obtain chemical reagents 

 which contained no lithium. We found it in 

 the distilled water and in nearly all the re- 

 agents. If there were none in any reagent, it 

 was only necessary to leave the reagent for a 

 little while in a glass vessel; after some time, 

 traces of lithium could be detected. The fol- 

 lowing experiment was made: When water 

 has been distilled from a platinum retort and 

 has been kept in a platinum flask, it leaves no 

 visible residue when 250 c.c. are evaporated in 

 a platinum dish; and the last drop, when 

 thus concentrated, does not show the spectrum 

 of lithium. But if water, prepared in the 

 same way, is kept for twenty-four hours in a 

 glass vessel, we find, on evaporation, a small 

 residue consisting chiefly of a sodium salt but 

 containing also a trace of lithium. 



It seemed to us essential to replace glass by 

 some other material. We have established 

 that it is just as dangerous to use quartz, the 

 material actually employed by Mr. Kamsay. 

 The commercial quartz vessels contain lith- 

 ium. Hydrofluoric acid containing no lithium 

 was allowed to act on the fragments of 

 an opaque quartz crucible and on a piece of 

 a transparent quartz tube. In the residue 

 there could be detected a notable quantity of 

 lithium, and there was a good deal more of it 

 in the transparent quartz than in the opaque 

 quartz. We then decided to use platinum 



The apparatus which we used consisted of 

 a cylindrical platinum vessel, Y.5 cm. in length 

 and 1.5 cm. outside diameter, placed hori- 

 zontally. At one end of the vessel was a 

 small vertical platinum tube through which 

 we could introduce the solution. The small 

 tube had a platinum cover which protected the 

 solution but which did not make a tight joint. 

 A glass tube, with a side-arm and a stop-cock, 

 was fastened on the outside of the platinum 

 tube. The solution was introduced into the 

 apparatus by means of a platinum siphon and 

 never came in contact for a moment with the 

 glass of the apparatus. 



The water and the acids necessary for the 

 experiment were redistilled in platinum retorts 

 and were kept in platinum bottles. We 



found that all these reagents originally con- 

 tained lithium, especially the sulphuric acid. 

 After purification lithium could not be de- 

 tected in the residues obtained by evaporating 

 80 c.c. of nitric acid, 25 c.c. of sulphuric acid, 

 25 c.c. of hydrofluoric acid and 250 c.c. of 

 water. 



As Mr. Eamsay himself points out, the so- 

 called chemically pure salts of copper contain 

 notable amounts of lithium. We have tried 

 different methods of purification : repeated pre- 

 cipitation by hydrogen sulphide; electrolytic 

 precipitation of copper; fractional crystalliza- 

 tion. We have finally made use of copper 

 sulphate which had been re-crystaUized a 

 number of times in a platinum dish, each 

 solution being made up with pure water. At 

 first this treatment is very effective; but it is 

 very difficult, if not impossible, to remove the 

 last traces of lithium. When we stopped 

 further purification we could just manage to 

 show the presence of lithium in the residue 

 resulting from treating 50 g. of copper salt; 

 but we could not detect its presence at all in 

 the residue resulting from the treatment of 

 2 g. of salt. 



The emanation was furnished by a solution 

 containing 0.19 g. of radium (0.25 g. EaClJ. 

 It was first condensed in a condenser im- 

 mersed in liquid air and then drawn over into 

 the apparatus. In order to know accurately 

 the amount of emanation introduced, we com- 

 pared the penetrating radiation of the ap- 

 paratus with that of a fiask containing a 

 known amount of radium. To do this we 

 used a specially-prepared condenser having 

 very large plates. 



Two exactly comparable experiments were 

 made. We introduced into the apparatus 

 about 7 c.c. of a pure copper sulphate solution ; 

 the liquid had a large free surface relatively 

 to its volume. The apparatus was sealed with 

 a blow-pipe. The emanation was introduced 

 in several lots; to ensure its dissolving, the 

 solution was shaken by tipping the apparatus 

 in its bath of melting ice. This operation was 

 repeated a number of times. The amounts of 

 metallic copper taken were 0.269 g. and 0.14 g. 

 The total amount of emanation introduced 

 was equivalent in both cases to the saturated 



