RICHARDS. — ATOMIC WEIGHT OP STRONTIUM. 379 



Siloe7\ — The preparation of pure silver has been repeatedly detailed. 

 The most elaborate method described iu the paper upon barium was 

 used in the present case.* A few improvements were introduced, 

 notably the purification of the sodic hydrate used for the reduction of 

 the argentic chloride by means of a strong galvanic current, instead of 

 by hydrogen sulphide. Little but iron was found in it, however. The 

 final crystals of electrolytic silver were usually fused upon pure sugar 

 charcoal or lime, in a reducing flame ; once however (for Analysis 10) 

 the crystals contained in a lime boat within a stout porcelain tube 

 were fused in a Sprengel vacuum by means of a Fletcher furnace. 

 Two holes bored through the furnace at right angles to the flame 

 entrance served to admit the tube. The heat was very gradually 

 applied, and after the silver had been melted all the apertures of the 

 furnace were closed and the tube was allowed to cool very slowly. A 

 wide glass tube set into the porcelain tube on one end served as a 

 convenient window for the observation of the fusion. 



Ammonic Carbonate. — Two varieties of ammonia carbonate were 

 used for the work just described. The first consisted of ordinary pure 

 " ammonic carbonate," which had been dissolved, treated with a 

 small amount of a pure strontium salt, and filtered. This treatment 

 undoubtedly removed any substance which could seriously interfere 

 with the preliminary purifications for which this ammonic carbonate 

 was used. For the final stages of the purification of the strontium 

 preparations, ammonic carbonate was made by saturating the purest 

 water in a platinum vessel with ammonia gas obtained by boiling the 

 pure strong ammonia of commerce, and then passing into this saturated 

 solution pure carbon dioxide. This latter gas was prepared by the 

 action of dilute nitric acid on marble ; it was purified by passing 

 through washing flasks containing water and a meter of glass tube 

 packed with moist beads. Upon delivering the gas into a Bunsen 

 flame, no trace of calcium could be detected spectroscopically. Both 

 gases were conducted into the solution through a platinum tube made 

 for the purpose. The resulting ammonic carbonate undoubtedly con- 

 tained more or less of the amines common in ordinary ammonia, but 

 it could not have contained a trace of non-volatile impurity capable 

 of contaminating the strontic carbonate for whose preparation it was 

 designed. 



Ammonic Oxalate. — This salt was made by neutralizing pure am- 

 monia water with pure oxalic acid, which had been still further 



* These Proceedings, XXIX. 64, 65. 



