974
WATER.—Analysis. Ammonia.
desirable that water which is to be used for drinkingand cooking, such as the water supply of a town,should contain in 100,000 parts more than 0-2 partand 002 part of these constituents respectively.Unpolluted upland surface waters contain, whentolerably free from peaty matter, from 0T to 0-2 partof carbon, and from 0 - 01 to 0-03 part of nitrogen.Water which has drained from a peat district isusually more or less brown in colour, and the car-bon is increased in a larger ratio than the nitrogen.Water from cultivated surfaces contains a largerproportion of organic matter, the organic carbonbeing usually from 0-25 to 0-3 part, and the organicnitrogen from 004 to 0-05. In shallow wellwaters the variation is excessive, and no rangecan be stated. In deep well waters the organicmatter is generally low, being usually about 0-06of carbon and 0 02 of nitrogen, or less. Springwaters much resemble deep well waters. In boththe character of the water is very constant, theconditions being practically unaffected by surfacechanges; whilst in the preceding class of shallowwells the reverse is the case, meteorological varia-tions and the incessant changes of surface pollutionat once affecting the water of the well. This rendersall such wells dangerous as sources of water supply,as there is no security that, even if the water shouldbe pure to-day, it may not be fatally pollutedto-morrow. The oxidising action which effectuallydestroyed a certain quantity of noxious matter maybe in great part suspended, or may be incapable ofdestroying a larger quantity; moreover, althoughdead organic matter may be oxidised, we do notknow that the same may be expected in regard tothe specific poisons of disease, which are organizedand living. Average sewage contains about 4 partsof organic carbon and 2 parts of organic nitrogen,but the range is very great.
The ratio of the organic carbon to the organicnitrogen is important, being as a rule high whenthe source of the organic matter is vegetable, andlow when it is animal. When organic matter ofvegetable origin is oxidised, it loses carbon morerapidly than nitrogen, so that the ratio is graduallyreduced; when of animal origin, it loses nitrogenmore rapidly than carbon, and the ratio thereforeincreases. In peaty waters it amounts to about 12
Q
on the average (i.e. — = 12), but may reach 20 or
even a larger amount; upland surface water con-taining but a small proportion of peat gives a ratioof 6 to 12; water from cultivated areas about0; shallow wells about 4, with a very wide range;deep wells and springs about 4, with a much narrowerrange; sea-water about 1-7 ; sewage about 1 to 3,with an average of 2.
Ammonia .—For the estimation of ammonia Ness-leii’s solution and a standard ammonic chloridesolution are required. For the former, dissolveG2-5 grams, of potassic iodide in 250 c.c. of distilledwater, and add gradually cold saturated solution ofcorrosive sublimate, until the red precipitate formed
on each addition just ceases to be redissolved onstirring. Then dissolve 150 grams, of solid potassichydrate in 150 c.c. of water, allow the solution tocool, add it to the first solution, make up with dis-tilled water to one litre, mix well, and allow tostand. A brown precipitate is deposited, the clearsolution being of a pale greenish yellow colour. Assoon as it is perfectly clear it is ready for use.
The standard solution of ammonic chloride ismade by dissolving 1-5735 gram, of the pure dry saltin water, and making up with water to one litre.This solution should be diluted to ten times itsvolume for use; 1 c.c. will then correspond to0-00005 gram, of ammonia (NII 3 ).
The estimation depends on the formation of ayellow or brownish coloration when Kessler’ssolution is added to a highly dilute ammoniacalsolution ; the intensity of the colour being estimatedby comparison with that produced by a knownquantity of ammonia. Add to about 50 c.c. of thewater to be examined, in a white glass cylinder about150 m.m. high and of about 70 c.c. capacity, about 1c.c. of Kessler’s solution, stir with a clean glass rod,and allow to stand for five minutes. If the colourthen produced, as seen standing on a white surface,is not deeper than that developed in a similarcylinder in which 0-1 c.c. of the standard ammonicchloride has been diluted with 50 c.c. of water freefrom ammonia and similarly treated with Kessler’ssolution, half a litre of the water should be takenfor the determination. Should the tint be deeper,less can be taken; but it is not desirable to employless than 20 c.c. Distilled water, as ordinarily used,is occasionally free from ammonia; but to obtain itas required above, it is generally necessary to rectifyit over freshly ignited sodic carbonate, of whichsufficient should be added to render it distinctlyalkaline. It is then distilled until, on testing thedistillate with Nessler’s solution as above, no traceof yellow is observed after standing for five minutes.The distillate which subsequently passes over is pre-served as “waterfree from ammonia.” The distilla-tion must not be continued to dryness.
If the sample is to be filtered and contains butlittle ammonia, the filter paper must not yield am-monia. If it is found to do so it must be steepedfor a day or two in water free from ammonia, thentransferred while wet direct to the funnel, and thefirst portion of the filtrate rejected. Washing withwater, even repeated, is usually insufficient. Unlessthe sample is very turbid it is not usually necessaryto filter it for the estimation of ammonia, althoughit is to be filtered for those of organic carbon andnitrogen and total solid matters.
A large glass tubulated retort should be used,connected by a short piece of caoutchouc tube withthe glass inner tube of a Liebig ’s condenser. Toinsure that the retort and condenser are free fromammonia, it is desirable to distil a little common ordistilled water with a little sodic carbonate until thedistillate is free from ammonia. In making a seriesof estimations on the same day it is only necessaryto do this at the beginning of the series, provided