EXPERIMENT STATION BULLETINS. 397 



eluding the stem below the stopcock should be filled with water while 

 the 10 c.c. burette should contain the sample which should likewise 

 fill the capillary stem below the stopcock. Both of these together with 

 the delivery tube should be in position in the rubber stopper. 



For carrying out the determination, 60 c.c. of a 30 per cent solution 

 of sodium nitrate are placed in the bottle, 15 c.c. of glacial acetic acid 

 added and the bottle wired on to the stopper, the gas escaping through 

 the side outlet of the three-way cock in the delivery tube. The stop- 

 cock of the cylinder is next opened, allowing water to run in and com- 

 pletely fill the bottle and that part of the delivery tube below the stop- 

 cock, thus displacing all air in the apparatus. The three-way cock in 

 the delivery tube can be completely closed by giving it a slight turn, 

 causing the gas which is being generated in the bottle to force the 

 liquid back into the cylinder. To insure complete removal of air from 

 the apparatus, the bottle and delivery tube are filled a second time with 

 liquid, the excess which escapes througli the side arm of the cock (K) 

 being caught in the flask (H). This stopcock is then again closed and 

 15-20 c.c. of gas allowed to accumulate in the bottle. After closing the 

 outlet to the cylinder, the three-way cock in the delivery tube is turned 

 so as to connect the bottle with the gas burette and the desired amount 

 of sample run in from the 10 c.c. burette. The sample should contain 

 about 5 mg. amino nitrogen per 10 ce. From 5-10 minutes are required 

 for the completion of the reaction, during which time from 80 to 100 

 c.c. of gas accumulate in the gas burette. After some practice the 

 amount can be so regulated by shaking the bottle that it becomes nearly 

 constant. At the end of the reaction the stopcock of the cylinder is 

 again opened and all the gas in the bottle and delivery tube replaced 

 by liquid. This stopcock is then closed, that in the delivery tube is 

 turned so as to permit the escape, through the side arm, of the gas from 

 the bottle while the three-way cock in the gas burette is completely 

 closed. The gas can be driven into the Hempel apparatus by raising 

 the leveling bulb. After shaking out the nitric oxide with the perman- 

 ganate solution the nitrogen is run back into the gas burette for meas- 

 urement. 



In preparing solutions for analysis care must be taken that all 

 ammonium compounds are removed and that the solution is not acid 

 with a strong mineral acid. Ammonium compounds react slowly^ 

 with nitrous acid causing high results while strong mineral acids react 

 too violently with sodium nitrite. Both of these difficulties can be over- 

 come in one operation by adding to a known quantity of the acid solu- 

 tion a slight excess of magnesium oxide and boiling off the free ammonia. 

 The solution can then be filtered and washed from the solid residue 

 and made up to a convenient volume for analysis. The slight alkalin- 

 ity of the solution after this treatment does not affect the determina- 

 tion. 



In calculating the percentage of nitrogen in the sample, it should 

 be remembered that, of the total volume of nitrogen obtained in the 

 gas burette, only one-half comes from the sample, the other half com- 

 ing from the nitrous acid. This fact adds considerably to the accuracy 



iConcerning the effect of ammonium compounds in results obtained by tliis method see tlie following 

 Kreusler, Vers. Stat. 31, 277: 294. (1885). Emmerling, Vers. Stat. 32, 440. (188G). Van Slyke. 

 Ber. d. Deutsch. Chem. Gesel. 43. 3170. (1910). 



