46 SEWAGE AND ITS PURIFICATION 



The organic N is always higher than the N as albuminoid 

 ammonia. Dr. McGowan proposes to call the difference the 

 *' X " nitrogen. This quantity is considerably lower in a good 

 effluent than in a raw sewage, showing that the organic matter 

 in the effluent is more easily broken up by the permanganate. 



With sewages and tank effluents the results thus easily and 

 rapidly obtained were useful, though they had a tendency 

 to be too low. But now and then, especially with good 

 effluents, the figures came out lower than the free ammonia. 

 To see whether the presence of nitrates or nitrites was the 

 cause, I made some experiments, of which the details are given 

 in the former edition, by adding these to an effluent containing 

 much ammonia, and found that a large excess of sulphuric 

 acid (4 c.c. as above) prevents loss of nitrogen by secondary 

 action. 



Accuracy is greatly affected by the fact that, in Nesslerizing, 

 the few c.c's of solution have to be multiplied by such a large 

 number to obtain the analytical figures. If time and quantity 

 allow, the Kjeldahl can be conducted in the ordinary way 

 with larger quantities — say, 100 c.c. of sewage, 200 or more 

 of effluent, with 10 c.c. of H2SO4, a globule of mercury, and 

 about J gramme of potassium sulphate, finally distilling with 

 pure soda and a little sodium sulphide into standard acid, and 

 titrating the loss of acidity.^ 



It is evident that among the unlimited number of organic 

 compounds that may be present in sewage there are some that 

 are very difficult to decompose, and Phelps^ finds this the 

 main cause of deficient organic nitrogen in Kjeldahls. He 

 records good results, usually within 0*05 part per 100,000, 

 by the following procedure, which has been adopted by the 

 Massachusetts Institute : 100 c.c. of the sample are heated 

 with 5 c.c. of sulphuric acid and about 0*1 gramme of copper 

 sulphate until perfectly colourless. He shows that stopping at 

 '*pale yellow" or "faint straw colour," as sometimes allowed, 

 is not sufficient. Condensing the vapours proved that a little 

 nitrogenous organic matter was volatilized, but the quantity 

 was negligible. In some cases the colourless residue still 

 contains some organic matter unresolved ; therefore he adds 

 in all cases small crystals of potassium permanganate until 

 there is a permanent colour. Experiments showed that while 



^ For details see Sutton's " Volumetric Analysis," i8g6, p. 85, 



■•^ Report of the Massachusetts Institute of Technology^ Boston, 1905, p. no. 



