107 



On consult iujr the literature it was fouud tliat iu 1877 Gatehouse pub- 

 lished a short note in the Chemical News giving the results of an experi- 

 ment on heating a mixture of these substances. He observed that the gas 

 was nitrogen, and from its volume he developed an equation to explain the 

 reaction. According to him each molecule of the oxide reacts with four 

 molecules of the nitrate ; producing six atoms of nitrogen and a molecule 

 of manganese nitrate. The presence of the last substance he does not 

 seem to have confirmed ; and as the temperature to which he heated the 

 mixture was above that at which manganese nitrate decomposes it ap- 

 peared doubtful that the final products were as he thought. The work be- 

 gun by Mr. Haworth has been continued by the writer and the nature of 

 the reaction mside more nearly complete. 



In our earlier experiments some difficulties were encountered and 

 some fa<'ts observed which modified the procedure in the later work. First, 

 it was fouud most difficult so to regulate the temperature that the reac- 

 tion would go smoothly and iu one direction. The action would proceed 

 at about 200° until nearly one-half of the gas had been evolved, and then 

 suddenly without apparent cause the thermometer would suddenly mount 

 to 300° or more and brown gases be evolved in such quantities that the 

 stopper and connections would be forced out with explosive violence. It 

 was thought at first that manganese nitrate was being formed in the 

 earlier stage of the reaction aud later was decomposing with evolution of 

 heat : but experiments with this substance showed that it decomposed in a 

 regular manner between 130° and 185°. Rut by heating for a time to 

 210°-220° and then cooling to 170" as the action proceeded it was found 

 possible to regulate the decomposition and get consistent results. 



It was also noted that after extracting the residue to determine the 

 amount of soluble material the aqueous solution was very strongly acid 

 with what appeared to be a nitrogen acid. The residue left on evaporation 

 consisted of unchanged ammonium nitrate, and at times of a trace of a 

 manganese compound. 



In the succeeding experiments the ai)paratus was so modified that the 

 gases could be i)assed through water to absorb the acid, and then collected 

 in a large gas burette made with litre cylinders and filled with dilute 

 alkali. About equal weights of ammonium nitrate and manganese dioxide 

 were placed in a distilling flask connected with the acid absorbing bottle 

 and the air in the apjiaratus replaced with nitrogen. The mixture was 

 heated to 170° and then connected to the gas burette. Afterwards the 



