OF ARTS AND SCIENCES. 207 



ways appeared to hold approximately the same amounts of gas. It is 

 almost, if not quite, impossible to prepare the oxide in this manner in 

 a state wholly free from solid impurities taken from the containing 

 vessel during the ignition of the purest possible zincic nitrate. The 

 effort was made in the preceding series of analyses to prepare samples 

 which must contain wholly different kinds of impurities. The fact that 

 these different samples contained almost equal amounts of gas shows 

 with reasonable certainty that the impurities are not responsible for 

 the occlusion. 



It is noticeable that the oxide obtained at a very low temperature, 

 which still contained traces of zincic nitrate, contained little or no 

 occluded gas (Experiments 8 and 30) ; also that six specimens which 

 had been suspected of partial reduction contained much less gas than 

 similar material free from this suspicion (Experiments 17, 24, 25, 

 34,35, and 36). 



NiCKELOus Oxide. 



The series of experiments with nickelous oxide led to results not 

 unlike those with zincic oxide. In this case sulphuric acid proved 

 unsatisfactory as a solvent, and hydrochloric acid was adopted. A 

 solution containing about twenty per cent of hydrochloric acid gas 

 was freed from air by continued boiling, rapidly cooled, and run into 

 the tube containing the oxide of nickel. On account of the very slow 

 action of the cold acid the tube was warmed after exhausting the air 

 as usual. The gas set free was measured as before. 



In order to prove the accuracy of the method a gram of zincic oxide 

 prepared from the carbonate was dissolved in hot hydrochloric acid in 

 precisely a similar way. No trace of gas was evolved during this 

 solution. 



It was thought unnecessary to make a series of experiments as 

 elaborate as that made with the zincic oxide. Nickelous nitrate was 

 evaporated to dryness in porcelain and ignited fifteen minutes in a 

 blast lamp and then a number of hours over a Buuseu burner in the 

 furnace. (Analysis 1, below.) 



A part of the nickelous oxide remaining was further ignited for two 

 hours in the furnace by means of the water blast. This was used for 

 the second analysis. A portion of the latter was ignited again in the 

 furnace at the highest temperature we could obtain, by the addition 

 of oxygen to the flame for about fifteen minutes, until the bottom of 

 the outer crucible as well as the platinum foil between the two cru- 



