THE TRANSFORMATION OF ELEMENTS 231 



than one dose of emanation is added. The additions are made in 

 a similar manner, at intervals of three or four days. In each 

 case the gases formed since the last addition are previously 

 pumped off as completely as possible and stored. The total 

 quantity of gas is analysed at the end of the experiment (see 

 below, (d) ), and the solution also analysed. 



(c) Method of analysis of the solutions. — In each case the copper 

 nitrate solution was transferred to a platinum crucible, and 

 saturated with sulphuretted hydrogen passed through a platinum 

 tube. The sulphide precipitate was separated by centrifugalisa- 

 tion in a silica tube, and the filtrate evaporated to dryness and 

 ignited. The residue was treated with a few drops of pure 

 water, and the solution, which should contain all the alkali salts, 

 was evaporated, ignited, weighed, and examined spectroscopically. 

 The whole of the analysis was carried out in either silica or 

 platinum vessels. 



In both experiments calcium and sodium (strong) were 

 present. Lithium was present in the " treated" case only. The 

 weight of the alkali residue was different in the two cases : that 

 from the " treated " solution was distinctly larger. This would 

 lead to the assumption that sodium is also a transmutation 

 product, but further evidence is necessary before stress can be 

 laid on these weight relations. The treatment of the pure water 

 in the third experiment has been indicated already. 



id) Method of analysis of the gaseous products. — The methods 

 employed for determining the amounts of the ordinary con- 

 stituents of a gas mixture were those devised by Ramsay himself. 

 They are applicable to quantities as small as one or two cubic 

 centimetres, the results in such cases being accurate to within 

 1 percent. The actual measurements are carried out in a gas 

 burette provided with a series of points sealed to the glass, 

 similar to those shown in fig. 3, Part II. p. 83, and the measure- 

 ments are made in a similar manner, the pressure exerted by the 

 gas being measured, from which its volume can be calculated. 

 (The calibrated gas volume is that from the tap to a mercury- 

 point surface.) 



In the gas-mixtures dealt with, the constituents are, as a rule, 

 only hydrogen, oxygen, the oxides of carbon, nitrogen, and inert 

 gases. The total volume is first measured. The gas is then 

 forced back through the inverted siphon into a test-tube and 

 allowed to remain for a few minutes in contact with moist 



