o 



OF ARTS AND SCIENCES. 6-; 



determinations were made in the air, but the results showed that, at 

 the temperature of volatilization, the iodide was rapidly oxidized, and 

 the determinations were afterwards made in an atmosphere of nitro- 

 gen. Many unsuccessful attempts were made to obtain pure nitrogen 

 by the well-known method described by Dr. Wolcott Gibbs. The 

 method consists in simply mixing together and warming solutions of 

 sodium nitrite, ammonium nitrate (or suljjhate) in excess, and potas- 

 sium dichromate in large excess, the speed of the reaction being regu- 

 lated by the amount of the dichromate added. It was found that red 

 fumes always appeared in the nitrogen evolved ; and that, while no 

 variations of proportions quite eliminated the fumes, still, when but 

 very little bichromate was used, the resulting nitrogen was compara- 

 tively pure. It was evident, as Dr. Gibbs pointed out, that the nitric 

 fumes were caused by chlorine present as an impurity in the materials 

 used, and the impurity was readily traced to the dichromate ; but, as 

 the purification of potassium dichromate is troublesome and expensive, 

 it was found more practicable to pass the impure nitrogen over red- 

 hot copper filings recently reduced, and then through drying-tubes. 

 The nitrogen thus purified gave satisfactory results. 



. Results. 



No. 1 17.25 



No. 2 17.90 



No. 3 17.63 



Mean 17.59 



Theory 17.33 



Greatest difference . . . 0.31 

 Mean difference . . . . 0.26 



It will be seen that the results do not agree very closely with the 

 theory, or with each otlier. While we do not presume that we have 

 obtained as accurate results as the method will allow, a ^evr trials care- 

 fully conducted and closely watched made it evident that very close 

 results cannot be reasonably expected from this method. An abso- 

 lutely constant temperature — indicated by the unchanging position 

 of the water surface in the delivery tube — it is practically impossible 

 to maintain. Furthermore, the slight resistance to be overcome by 

 the air in rushing out of the delivery tube is often enough to retain 



