WATER.—Analysis. Nituates and Nitrites.
975
that each distillation is continued until ammoniaceases to come over. The residue is removed tomake way for another estimation by means of aclean glass siphon.
The retort being free from ammonia, the measuredquantity of the water is introduced, together withabout 1 gram, of recently ignited sodic carbonate,and, if necessary, enough water free from ammoniato make up the volume to about 400 c.c., and thedistillation is rapidly proceeded with; the distillatebeing collected in successive portions of about 50 c.c.each in small glass cylinders previously described.It is best to begin by examining the second of theseas soon as it is collected; add to it about 1 c.c. ofNessler’s solution, stir, and allow it to stand forfive minutes. Then observe the colour produced,as seen standing on a white surface, and measurefrom a narrow burette into a similar cylinder aquantity of standard ammonic chloride which youconsider will give, when tested, an equal depth ofcolour. Add to this about 50 c.c. of water freefrom ammonia and then 1 c.c. of Nessler’s solution,and allow it also to stand for five minutes. If thecontents of the two cylinders display the same tint,they contain equal quantities of ammonia; but ifnot, fresh trial cylinders with a greater or lessquantity of ammonia must be prepared until equalityis gained. Ammonia solution must not be addedafter the Nessler’s solution, or a precipitate will beproduced.
Nessler’s solution may be added to the secondand subsequent fractions of distillate as soon as eachis distilled over, and when one is obtained free fromammonia the distillation may be stopped. If theammonia in the second fraction does not exceed thatin 0-4 c.c. of the standard solution, Nessler’s solu-tion may be added to the first fraction, and theammonia in it estimated as above; but if it exceedsthat, the first fraction should be measured and analiquot part taken and diluted to about 50 c.c. withwater free from ammonia, and the ammonia deter-mined in this part, as the undiluted fractions wouldprobably give a colour too dark for convenient esti-mation. When the estimation is thus made in apart it is as well to take another portion and a freshtrial cylinder, adding the Nessler’s solution to bothat the same time, so that the tints may be comparedwhen they have been standing for the same lengthof time. If it is known that a large quantity ofammonia is present, time is saved by at once takingas a first fraction about 100 c.c. and diluting a smallmeasured part of this. The volumes of standardammonic chloride solution corresponding to thevarious fractions are added together; and if 500 c.c.of the sample were used, the total number of c.c. ofthe standard solution divided by 100 will express theammonia (N1I 3 ) present in parts per 100,000.
The ammonia in water is not always of animalorigin. In upland waters the average is about 0-002part per 100,000; in that from cultivated areas about0-005, with a considerable range; in water from shal-low wells all proportions from nil to 2-5 parts have ■been noticed; in deep well waters the average is |
j about 0-001 part, but sometimes in excellent sam-j pies amounting to 0-1 part, or even somewhatmore; in spring waters the average is only 0-001.Sewage usually contains from 2 to 6 parts, with anaverage of about 5. The occurrence of a consider-able quantity of ammonia usually coincides withmuch organic matter and deficient oxidation; thatfound in water from some very deep wells probablyresults from decomposition of nitrates.
Nitrogen as Nitrates and Nitrites .—This is deter-mined by a modification of a process devised byWalter Crum for the refraction of nitre. Theresidue from the determination of the total solidmatters is treated with 20 to 30 c.e. of hot distilledwater, and when the whole has been well wettedwith the water, enough of a saturated solution ofsulphate of silver is added to precipitate the wholeof the chlorine present, and the dish is then heatedon a steam bath for some minutes. The sulphate,which is but sparingly soluble, is prepared by actingupon well-washed carbonate of silver with a smallexcess of sulphuric acid, carefully neutralizing thefree acid with caustic potash, and washing the sul-phate with cold distilled water until, on evaporatingabout 100 c.c. of the solution to dryness, and exam-ining the residue according to the method about tobe described, it yields no evidence of nitrates. Thesilver is, of course, recovered from these washings.If excess of sulphate of silver has been added, theresidue usually becomes dark brown in colour. Itshould be tested by adding a drop of the sulphate;and if there is still a precipitate, a further quantitymust be added until all the chlorine is thrown down.As soon as this is accomplished, and the contents ofthe dish evaporated to a very small bulk—some 5 or10 c.c.—filter through a very small filter of Swedish or German paper, and wash several times with thesmallest possible quantity of hot distilled water.The better to extract the nitrates from the dish, it iswell to rub the residue off the surface with the endof the finger, protected by a caoutchouc finger-stall.The dish is readily cleaned by plunging it into dilutesulphuric acid and placing in contact with it a pieceof zinS. By this means the silver is reduced to themetallic state, and most of it may be easily detached;the dish is then washed, and the remainder of themetallic silver removed by a bath of warm dilutenitric acid.
The filtrate is evaporated in a very small beakerto dryness, or nearly so, over a steam bath. Thelittle apparatus shown in Fig. 9 is then placed in anordinary porcelain mercury trough, and filled withmercury just above the stop-cock. The total length
of the tube is 210 m.m. and the internal diameterabout 15 m.m. The concentrated solution, obtainedas above, is then poured into the cup at the top ofthe tube; and the stop-cock being gently opened, it isallowed to flow down into the lower tube without