120 
2. This residue was then heated for some time to redness 
to destroy and free it from any organic bodies and again 
weighed. 
3. The remaining mineral matters were then submitted 
to analysis by separation of the different salts by means of 
alcohol of different strengths to corroborate the results of 
the more exhaustive analysis. 
4. 5 cubic centimetres of the solution were mixed with 
245c.c. of pure water, free from ammonia, distilled and the 
free ammonia determined by means of “ nesslerizing” to shew 
the proportion of ammonia iri combination with carbonic acid. 
Pure carbonate of soda solution and pure water were added 
to the liquid remaining in the retort and it was again dis- 
tilled till the distillate ceased to shew any coloration by the 
nessler test, aud the ammonia from this distillate estimated 
by nesslerizing. This operation would give the proportion of 
ammonia which, although in the analysis of potable water 
would be termed “ free ammonia ” would in this case be in 
combination with hydrochloric acid. More pure water and 
permanganate of potassium were added to the solution, again 
remaining in the retort, and the liquid again boiled and the 
distillate collected with the same precautions as before, and 
the ammonia contained in it estimated. This should give 
an approximate idea of the nitrogen which is contained in 
the water, not as ammonia, but presumably in combination 
in some organic compound. 
5. 100 c.c. of the sample were heated with a little 
free sulphuric acid, and a standard solution of potassium 
permanganate added till the liquid ceased to decolorize it, to 
find the amount* of oxygen required to oxidise any matters 
capable of oxidation by such means. 
The following determinations were made in the original 
water : — 
Chlorine . — This was estimated by standard nitrate of 
silver solution. 
