EXPRESSION OF RESULTS.

The results of chemical analyses shall be expressed in parts per million, which in most analyses is practically equivalent to milligrams per liter. In some laboratories other forms of expression have been used. Results expressed in parts per 100,000 or in grains per gallon may be transformed to parts per million, or conversely, by the use of the following table:

Table 4.—Factors for transforming results of analyses.
Unit.	Equivalent.
	Grains per U.S. gallon.	Grains per Imperial gallon.	Parts per 100,000.	Parts per million.
1 grain per U. S. gallon	1.000	1.20	1.71	17.1
1 grain per Imperial gallon	.835	1.00	1.43	14.3
1 part per 100,000	.585	.70	1.00	10.0
1 part per million	.058	.07	.10	1.0

The following general rules shall govern the use of significant figures in the expression of results:

1. If the results show quantities greater than 10 parts per million use no decimals; record only whole numbers. If the quantities reach hundreds and thousands of parts record only two significant figures.

2. If the results are between 1 and 10 parts do not retain more than one decimal place.

3. If the results are between 0.1 and 1 part do not retain more than two decimal places.

4. Estimates of ammonia, albuminoid, and nitrite nitrogen alone justify the use of three decimals.

5. If the results of analyses are tabulated ciphers should not be added at the right of the decimal point to make the column uniform.

Nitrogenous organic matter passes through several intermediate compounds during its natural decomposition, and that which does not gasify ultimately forms nitrate. Nitrogen in organic matter is determined by the Kjeldahl process.^[13][14][58] An indication of the amount present is obtained by the albuminoid nitrogen determination.^{[14][15][67][106][107]} It has not been found possible to differentiate the nitrogen in the organic matter that readily decomposes from that in stable or non-putrescible compounds. Decomposition of organic matter produces nitrogen combined in ammonia, which is the first step between nitrogenous organic matter and the completely mineralized nitrate. Ammonia nitrogen may be determined by distillation and Nesslerization or by direct Nesslerization of the clarified sample. The next step is oxidation to nitrite, and the final step, oxidation to nitrate. It is recommended that all forms of nitrogen be reported as the element nitrogen (N).

AMMONIA NITROGEN.

There are two methods for estimating ammonia nitrogen—distillation and direct Nesslerization. Distillation is recommended for most waters and direct Nesslerization is recommended for sewages, sewage effluents, and highly polluted surface waters.

DETERMINATION BY DISTILLATION.^{[38][68b][111][121]}

Procedure.—Use a metal or a glass flask connected with a condenser so that the distillate may drop from the condenser tube directly into a Nessler tube or a flask. Free the apparatus from ammonia by boiling distilled water in it until the distillate shows no trace of ammonia. After this has been done empty the distilling flask and measure into it 500 cc. of the sample, or a smaller portion diluted to 500 cc. with ammonia-free water. If the sample is acid or if the presence of urea is suspected add about 0.5 gram of sodium carbonate before distillation. Omit this if possible as it tends to increase “bumping.” Apply heat so that the distillation may proceed at the rate of not more than 10 cc. nor less than 6 cc. per minute. Collect the distillate in four Nessler tubes, 50 cc. to each tube, or if the nitrogen is high in a 200 cc. graduated flask. These receptacles contain the ammonia nitrogen to be measured as hereafter described.

Use Nessler tubes of such diameter that the graduation mark is between 20 and 25 cm. above the bottom and of such uniformity of diameter that the distance from the bottom to the graduation mark of the longest tube shall not exceed that of the shortest tube by more than 6 mm. The tubes must be of clear white glass with polished bottoms.

MEASUREMENT OF AMMONIA NITROGEN.

The amount of ammonia in the distillates may be measured either by (1) comparison of the Nesslerized distillates with Nesslerized solutions containing known quantities of nitrogen as ammonium chloride, or by (2) comparison of the Nesslerized distillates with permanent standard solutions in which the colors of Nesslerized standard ammonia solutions are duplicated by solutions of platinum and cobalt chlorides.

Comparison with ammonia standards.

Reagents.—1. Ammonia-free water.

2. Standard ammonium chloride solution. Dissolve 3.82 grams of ammonium chloride in ammonia-free water and dilute to 1 liter; dilute 10 cc. of this to 1 liter with ammonia-free water. One cc. equals 0.00001 gram of nitrogen.

3. Nessler reagent.^[8] Dissolve 50 grams of potassium iodide in a minimum quantity of cold water. Add a saturated solution of mercuric chloride until a slight precipitate persists permanently. Add 400 cc. of 50 per cent solution of potassium hydroxide, made by dissolving the potassium hydroxide and allowing it to clarify by sedimentation before using. Dilute to 1 liter, allow to settle, and decant. This solution should give the required color with ammonia within five minutes after addition and should not produce a precipitate with small amounts of ammonia within two hours.

Procedure.—Prepare a series of 16 Nessler tubes containing the following amounts of the standard ammonium chloride solution, diluted to 50 cc. with ammonia-free water, namely: 0.0, 0.1, 0.3, 0.5, 0.7, 1.0, 1.4, 1.7, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, and 6.0 cc. These solutions will contain 0.00001 gram of nitrogen for each cubic centimeter of the standard solution.

Nesslerize the standards and the distillates by adding approximately 1 cc. of Nessler reagent to each tube. Do not stir the contents of the tubes. The temperature of the tubes should be practically the same as that of the standards; otherwise the colors will not be directly comparable.^[45] Allow the tubes to stand at least 10 minutes after Nesslerizing. Compare the color produced in the tubes with that in the standards by looking vertically downward through them at a white or mirrored surface placed at an angle in front of a window so as to reflect the light upward.

If the color obtained by Nesslerizing the distillates is greater than that of the darkest tube of the standards, mix the contents of the tube thoroughly, pour out half of the liquid, and dilute the remainder to the original volume with ammonia-free water; then make the color comparison and multiply the result by two. If the color is still too dark after pouring out half the liquid, repeat this process of division until a reading can be made. The process of dilution may be shortened by mixing together the distillates from one sample before making the comparison and comparing an aliquot portion with the standards.

After the readings have been recorded add the results obtained by Nesslerizing each portion of the entire distillate. If 500 cc. of the sample is distilled this sum, expressed in cubic centimeters and multiplied by 0.02, will give the number of parts per million of ammonia nitrogen in the sample. If x cc. of sample is used multiply the sum of the readings by 10/x.

If the ammonia is known to be high the distillate may be collected in 200 cc. flasks and an aliquot part Nesslerized.

Comparison with permanent standards.^[62][65]

Reagents.—Platinum solution. Dissolve 2.00 grams of potassium platinic chloride (PtCl₄.2KCl) in a small amount of distilled water, add 100 cc. of strong hydrochloric acid, and dilute to 1 liter.

Cobalt solution. Dissolve 12 grams of cobaltous chloride (CoCl₂.6H₂O) in distilled water, add 100 cc. of strong hydrochloric acid, and dilute to 1 liter.

Prepare standards by putting various amounts of these two solutions into Nessler tubes and diluting to the 50 cc. mark with distilled water as indicated in Table 5. These standards may be kept for several months if protected from dust.

Table 5.—Preparation of permanent standards for the determination of Ammonia.
Value in standard ammonium chloride.	Solution of platinum.	Solution of cobalt.
cc.	cc.	cc.
0.0	1.2	0.0
.1	1.8	.0
.2	2.8	.0
.4	4.7	.1
.7	5.9	.2
1.0	7.7	.5
1.4	9.9	1.1
1.7	11.4	1.7
2.0	12.7	2.2
2.5	15.0	3.3
3.0	17.3	4.5
3.5	19.0	5.7
4.0	19.7	7.1
4.5	19.9	8.7
5.0	20.0	10.4
6.0	20.0	15.0
7.0	20.0	22.0

The amounts in Table 5 are approximate, and the actual amount necessary will differ with the character of the Nessler solution, the color sensitiveness of the analyst’s eye, and other conditions. The final test of the standard is best obtained by comparing it with Nesslerized standards and modifying the tint accordingly. Such comparison should be made for each new batch of Nessler solution and should be checked by each analyst.

Procedure.—In comparison with permanent standards, Nesslerize the distillates in the manner above described and compare the resulting colors at the end of about 10 minutes with the permanent standards. The method of calculating results is precisely the same as with the ammonia standards.

MODIFICATION FOR SEWAGE.

Ammonia nitrogen and albuminoid nitrogen in sewages, soils, and other materials of high nitrogen content may be satisfactorily determined by diluting the sample with ammonia-free distilled water and proceeding as described in the preceding sections, but it is permissible to distill with steam.^[40]

Procedure.—Use a 200 cc. long-necked Kjeldahl flask connected with a condenser so that the distillate may drop from the condenser tube directly into a Nessler tube or a flask. Connect the Kjeldahl flask with a steam generator by a tube reaching almost to the bottom of the flask.

After the apparatus is freed from ammonia put the sample to be tested into the flask. Use 10 to 100 cc. of the sample according to its ammonia content. Pass ammonia-free steam through the liquid in the Kjeldahl flask and collect the distillate in the usual way. It is usually convenient to collect the distillate in a 200 cc. flask and to take an aliquot part of it for Nesslerization. Compare with standards and calculate the nitrogen content in the usual manner.

This method has the advantage, when the sample is treated with an alkaline solution of potassium permanganate, of avoiding bumping, permitting the assay of solid matter, and yielding the ammonia more rapidly than by the ordinary process of distillation.

DETERMINATION BY DIRECT NESSLERIZATION.^[21][75]

Reagents.—

1. Ten per cent solution of copper sulfate (CuSO₄.5H₂O).

2.

Ten per cent solution of lead acetate (Pb(C₂H₃O₂)₂.3H₂O).

3.

Fifty per cent solution of sodium hydroxide (NaOH) or potassium hydroxide (KOH).

Procedure.—To 50 cc. of the sample to be tested, diluted if necessary with an equal volume of ammonia-free water, in a short tube, add a few drops of the copper sulfate solution. After thoroughly mixing, add 1 cc. of the alkali hydroxide solution and again thoroughly mix. Allow the tube to stand for a few minutes, when a heavy precipitate should fall to the bottom, leaving a colorless supernatant liquid. Nesslerize an aliquot part. Compare with standards and compute the ammonia nitrogen in the same manner as in the distillation procedure.

Samples containing hydrogen sulfide may require the use of lead acetate in addition to the copper sulfate. Some samples may require a few trials before the right combination of the three solutions to bring about the best results can be found.

Instead of adding copper sulfate to sewages of high magnesium content satisfactory clarification of the sample can be obtained by mixing it with the alkali hydroxide alone.^[54]

The addition of an alkaline permanganate solution to liquids containing nitrogenous organic matter causes the formation of ammonia, which can be distilled and determined by Nesslerization of the distillate. The nitrogen of the ammonia, thus obtained, is called albuminoid nitrogen. As the ratio of nitrogenous organic matter to the ammonia obtained by distillation is decidedly variable^[6][30][75] in sewages and other substances containing much nitrogenous organic matter albuminoid nitrogen results on such substances are less accurate^[29] than organic (Kjeldahl) nitrogen. Therefore in sewage work, including analysis of influents and effluents of purification plants and the water of highly polluted streams, it is recommended that determinations of organic nitrogen be substituted for determinations of albuminoid nitrogen. For ground waters and surface waters containing but little pollution, the albuminoid nitrogen is approximately one-half the organic nitrogen; accordingly the continuance of albuminoid nitrogen determinations for this class of work is approved.

Reagents.—Alkaline potassium permanganate. Pour 1,200 cc. of distilled water into a porcelain dish holding 2,500 cc., boil 10 minutes, and turn off the gas. Add 16 grams of C. P. potassium permanganate and stir until solution is complete. Then add 800 cc. of 50 per cent clarified solution of potassium hydroxide or an equivalent amount of sodium hydroxide and enough distilled water to fill the dish. Boil down to 2,000 cc. Test this solution for ammonia by making a blank determination. Correct determinations by the amount of this blank.

Procedure.—After the collection of the distillate for ammonia nitrogen described on page 15 add 50 cc. (or more if necessary to insure the complete oxidation of the organic matter) of alkaline potassium permanganate and continue the distillation until at least four portions, and preferably five portions, of 50 cc. each, of distillate have been collected in separate tubes. Determine the albuminoid nitrogen in the distillate by Nesslerization. If the albuminoid nitrogen is known to be high it is convenient to collect the distillate in a 200 cc. flask and to Nesslerize an aliquot part of it.

Dissolved albuminoid nitrogen may be determined in a sample from which suspended matter has been removed by filtration either through filter paper or through a Berkefeld filter. Suspended albuminoid nitrogen is the difference between the total and the dissolved albuminoid nitrogen.

ORGANIC NITROGEN.^{[24b][69][71][76][84]}

Procedure for water.—Boil 500 cc. of the sample in a round-bottomed flask to remove ammonia nitrogen. This usually causes the loss of 200 cc. of the sample, which may be collected for the determination of ammonia nitrogen. Add 5 cc. of nitrogen-free concentrated sulfuric acid and a small piece of ignited pumice. Mix by shaking and place over a flame under a hood. Digest until copious fumes of sulfuric acid are given off and the liquid finally becomes colorless or pale straw color. Remove from the flame, and add potassium permanganate crystals in small portions until a heavy green precipitate persists in the liquid. Cool. Dilute to about 300 cc. with ammonia-free water. Make alkaline with 10 per cent ammonia-free sodium hydroxide. Distill the ammonia, collect the distillate in Nessler tubes, Nesslerize, and compare with standards as described (pp. 16–18).

First procedure for sewage^[76].—Distill the ammonia nitrogen directly from 100 cc. or less of the sample, diluted to 500 cc. with nitrogen-free water. Collect the distillate and determine the ammonia nitrogen in it. Add 5 cc. of nitrogen-free sulfuric acid and 1 cc. of 10 per cent nitrogen-free copper sulfate, and digest the liquid for half an hour after it has become colorless or pale straw color. Add 0.5 gram of potassium permanganate crystals to the hot acid solution, and dilute to 500 cc. with ammonia-free water. Dilute 10 cc. or more of this liquid, in a Kjeldahl distilling flask, to about 300 cc. with ammonia-free water. Make alkaline with 10 per cent sodium hydroxide, distill, and Nesslerize. With some samples direct Nesslerization may be used. (See p. 19.)

In this determination care must be taken to digest thoroughly, to add potassium permanganate to the point of precipitation, to sample carefully after dilution, and to add enough sodium hydroxide to insure the separation of the ammonia from the precipitated manganese hydroxide. Potassium permanganate should not be added during digestion because it causes loss of nitrogen.

Second procedure for sewage.—Omit the separation of ammonia nitrogen and determine the ammonia nitrogen and organic nitrogen together. Determine the ammonia nitrogen in a separate sample by direct Nesslerization as described on page 19. The organic nitrogen is equal to the difference.

NITRITE NITROGEN.^{[51][63a][64][94c][108]}

Reagents.—1. Sulfanilic acid solution. Dissolve 8.00 grams of the purest sulfanilic acid in 1,000 cc. of 5 N acetic acid (sp. gr. 1.041) or in 1,000 cc. of water containing 50 cc. of concentrated hydrochloric acid. This is practically a saturated solution.

2. a-naphthylamine acetate or chloride solution. Dissolve 5.00 grams solid a-naphthylamine in 1,000 cc. of 5 N acetic acid or in 1,000 cc. of water containing 8 cc. of concentrated hydrochloric acid. Filter the solution through washed absorbent cotton or an alundum filter.

3. Sodium nitrite stock solution. Dissolve 1.1 gram silver nitrite in nitrite-free water; precipitate the silver with sodium chloride solution and dilute the whole to 1 liter.

4. Standard sodium nitrite solution. Dilute 100 cc. of solution 3 to 1 liter, then dilute 50 cc. of this solution to 1 liter with sterilized nitrite-free water, add 1 cc. of chloroform, and preserve in a sterilized bottle. One cc. = 0.0005 mg. nitrogen.

5. Fuchsine solution. 0.1 gram per liter.

Procedure.—Place in a standard Nessler tube 50 cc. of the sample, decolorized if necessary with nitrite-free aluminium hydroxide (see p. 42) or a smaller amount diluted to 50 cc. At the same time prepare in Nessler tubes a set of standards, by diluting to 50 cc. with nitrite-free water, various amounts of the standard nitrite solution. The following amounts of standard solution are suggested: 0.0, 0.1, 0.2, 0.4, 0.7, 1.0, 1.4, 1.7, 2.0, and 2.5 cc. Add 1 cc. of the sulfanilic acid solution and 1 cc. of the a-naphthylamine acetate or hydrochloride solution to the sample and to each standard. Mix thoroughly and allow to stand 10 minutes; then compare the sample with the standards. Do not allow the sample to stand more than one-half hour before making the comparison. If the color of the sample is deeper than that of the highest standard repeat the test on a diluted sample. If 50 cc. of the sample is used 0.01 times the number of cc. of the standard matched equals parts per million of nitrite nitrogen. Satisfactory results can be obtained by using either hydrochloric or acetic acid in preparing the test solutions, but the speed of the reaction is more rapid if acetic acid is used.^[112]

Permanent standards may be prepared by matching the nitrite standards with dilutions of the fuchsine solution. Fuchsine standards have been found to be sufficiently accurate for waters high in nitrite and for sewage. The standards should be checked once a month and kept out of bright sunlight.

NITRATE NITROGEN.^{[16][36][90][100]}

Two methods are recommended for the determination of nitrate nitrogen in water, sewage, and sewage effluents.

PHENOLDISULFONIC ACID METHOD.^[1][5][32]

Reagents.—1. Phenoldisulfonic acid. Dissolve 25 grams of pure white phenol in 150 cc. of pure concentrated sulfuric acid. Add 75 cc. of fuming sulfuric acid (15 per cent SO₃), stir well, and heat for 2 hours at about 100°C.

2. Potassium hydroxide solution. Prepare an approximately 12 N solution, 10 cc. of which will neutralize about 4 cc. of the phenoldisulfonic acid.

3. Standard nitrate solution. Dissolve 0.72 gram of pure recrystallized potassium nitrate in 1 liter of distilled water. Evaporate cautiously to dryness 10 cc. of the solution on the water bath. Moisten residue quickly and thoroughly with 2 cc. of phenoldisulfonic acid and dilute to 1 liter. This is the standard solution, 1 cc. of which equals 0.001 mg. of nitrate nitrogen.

4. Standard silver sulfate solution. Dissolve 4.4 grams of silver sulfate free from nitrate in 1 liter of water. One cc. of this solution is equal to 1 mg. of chloride.

Procedure.—The alkalinity, chloride, and nitrite content, and color of the sample must first be determined. If the sample is highly colored decolorize it with freshly precipitated aluminium hydroxide. Measure into an evaporating dish 100 cc. of the sample, or if nitrate is very high such volume as will contain about 0.01 mg. of nitrate nitrogen. Add sufficient N/50 sulfuric acid nearly to neutralize the alkalinity. Then add sufficient standard silver sulfate to precipitate all but about 0.1 mg. of chloride. The removal of chloride may be omitted if the sample contains less than 30 parts per million of chloride. Heat the mixture to boiling, add a little aluminium hydroxide, stir, filter, and wash with small amounts of hot water. Evaporate the filtrate to dryness, and add 2 cc. of the phenoldisulfonic acid, rubbing with a glass rod to insure intimate contact. If the residue becomes packed or appears vitreous because of the presence of much iron, heat the dish on the water bath for a few minutes. Dilute the mixture with distilled water, and add slowly a strong solution of potassium hydroxide or ammonium hydroxide until the maximum color is developed. Transfer the solution to a Nessler tube, filtering if necessary. If nitrate is present a yellow color will be formed. Compare the color with that of standards^[52][55] made by adding 2 cc. of strong potassium hydroxide or ammonium hydroxide to various amounts of standard nitrate solution and diluting them to 50 cc. in Nessler tubes. The following amounts of standard nitrate solution are suggested: 0, 0.5, 1.0, 1.5, 2.0, 4.0, 6.0, 8.0, 10.0, 15.0, 20.0, and 40.0 cc. These standards may be kept several weeks without deterioration. If 100 cc. of water is used the number of cubic centimeters of the standard multiplied by 0.01 is equal to parts per million of nitrate nitrogen.

Standards that will remain permanent for several years if stored in the dark may be prepared from tripotassium nitrophenoldisulfonate.^[5]

If nitrite nitrogen is present in excess of 1 part per million it should be oxidized by heating the samples a few minutes with a few drops of hydrogen peroxide free from nitrate repeatedly added^[95] or by adding dilute potassium permanganate in the cold until a faint pink coloration appears; the nitrogen equivalent of the nitrite thus oxidized to nitrate is then subtracted from the final nitrate nitrogen reading.

REDUCTION METHOD.^[2][46]

Reagents.—1. Sodium or potassium hydroxide solution. Dissolve 250 grams of the hydroxide in 1.25 liters of distilled water. Add several strips of aluminium foil and allow the evolution of hydrogen to continue over night. Concentrate the solution to 1 liter by boiling.

2. Aluminium foil. Use strips of pure aluminium about 10 cm. long, 6 mm. wide, and 0.33 mm. thick and weighing about 0.5 gram.

Procedure.—To 100 cc. of the sample in a 300 cc. casserole add 2 cc. of the hydroxide solution and concentrate by boiling to about 20 cc. Pour the contents of the casserole into a test tube about 16 cm. long and 3 cm. in diameter, or of approximately 100 cc. capacity. Rinse the casserole several times with nitrogen-free water and add the rinse water to the liquid already in the tube, thus making the contents of the tube approximately 75 cc. Add a strip of aluminium foil. Close the tube by means of a rubber stopper through which passes a bent glass tube about 5 mm. in diameter. Put the shorter arm of the tube flush with the lower side of the rubber stopper and let the longer arm extend below the surface of distilled water in another test tube. This apparatus serves as a trap through which the evolved hydrogen escapes freely. The small amount of ammonia escaping into the trap may be neglected. Allow the action to proceed for a minimum period of four hours or over night. Pour the contents of the tube into a distilling flask, dilute with 250 cc. of ammonia-free water, distill, collect the distillate in Nessler tubes, and Nesslerize. If the nitrate content is high collect the distillate in a 200 cc. flask and Nesslerize an aliquot part. If the supernatant liquid in the reduction tube is clear and colorless the solution may be diluted to a definite volume and an aliquot part Nesslerized without distillation.

TOTAL NITROGEN.^[93]

In sewage work it is frequently of assistance to know the total nitrogen content. This is ordinarily computed by adding together the organic, ammonia, nitrite, and nitrate nitrogen, each of which is determined as already described.

OXYGEN CONSUMED.^{[24][67][84a][85][94f][101][102]}

Oxygen consumed means the oxygen that the oxidizable compounds of sewage and water consume when treated in an acid solution with potassium permanganate. The expression is synonymous with oxygen required, oxygen absorbed, and oxygen-consuming capacity. It should not be confused with biochemical oxygen demand.

As the carbon, not the nitrogen, in organic matter is oxidized by potassium permanganate, oxygen consumed is considered by some an indication of the amount of carbonaceous organic matter present. The determination indicates, however, only part of the carbon, the proportion varying in different samples because the carbon in nitrogenous matter is not so readily oxidized as that in carbonaceous organic matter. Furthermore, it does not directly differentiate the carbon present in unstable organic matter from that in fairly stable organic matter, such as is sometimes referred to as residual humus matter. As nitrite nitrogen, ferrous iron, sulfide, and other oxidizable mineral substances reduce potassium permanganate, corrections for them should be made in the determination.

RECOMMENDED METHOD.

Reagents.—1. Dilute sulfuric acid. Dilute 1 part of concentrated sulfuric acid with 3 parts of distilled water and free the solution from oxidizable matter by adding potassium permanganate until a faint pink color persists after the solution has stood several hours.

2. Standard ammonium oxalate. Dissolve 0.888 gram of the pure salt in 1 liter of distilled water. One cc. is equivalent to 0.1 mg. of oxygen. An equivalent quantity of oxalic acid or sodium oxalate may be used.

3. Standard potassium permanganate. Dissolve 0.4 gram of the crystallized salt in 1 liter of distilled water. Add 10 cc. of the dilute sulfuric acid and 10 cc. of this solution of potassium permanganate to 100 cc. of distilled water, and digest 30 minutes. Add 10 cc. of the ammonium oxalate solution, and then add potassium permanganate till a pink coloration appears. This destroys the oxygen-consuming capacity of the water used. Now add another 10 cc. of ammonium oxalate solution and titrate with potassium permanganate. Adjust the potassium permanganate solution so that 1 cc. is equivalent to 1 cc. of ammonium oxalate solution or 0.1 mg. of available oxygen.

Acid digestion.—Place in a flask 100 cc. of the water, or, if the water is of high organic content, a smaller portion diluted to 100 cc. Add 10 cc. of sulfuric acid solution and 10 cc. of standard potassium permanganate and digest the liquid exactly 30 minutes in a bath of boiling water the level of which is kept above the level of the contents of the flask.^[70][71a] If the quantity of permanganate is insufficient for complete oxidation repeat the digestion with a larger quantity; at least 5 cc. excess of the standard permanganate should be present when the ammonium oxalate solution is added. Remove the flask, add 10 cc. of the ammonium oxalate solution, and titrate with the standard permanganate until a faint but distinct color is obtained. If 100 cc. of water is used the number of cubic centimeters of potassium permanganate solution in excess of the number of cubic centimeters of ammonium oxalate solution is equal to parts per million of oxygen consumed.

If oxidizable mineral substances, such as ferrous iron, sulfide, or nitrite, are present in the sample corrections should be applied as accurately as possible by suitable procedures. Direct titration of the acidified sample in the cold, using a three-minute period of digestion, serves this purpose quite well for polluted surface waters and fairly well for purified sewage effluents. Few raw sewages containing no trade wastes need such a correction, but raw sewages containing “pickling” liquors do need it. If the sample contains both oxidizable mineral compounds and gaseous organic substances the latter should be driven off by heat and the sample allowed to cool before applying this test for the correction factor. If such corrections are made the fact should be stated with the amount of correction.

Period and temperature of digestion.—As the practice in regard to the period and temperature of digestion has varied widely it is difficult to compare the results obtained at one laboratory with those obtained at another. None of the methods gives absolute results. They are all relative^[26][29][57] at best. Digesting 30 minutes at the boiling temperature is herein designated the recommended method. If samples are analyzed by any other method the method should be noted, and, representative results by the standard method should be placed on record for purposes of comparison.

OTHER METHODS.

Additional reagents.—1. Potassium iodide solution. Ten per cent solution, free from iodate.

2. Standard sodium thiosulfate. Dissolve 1.0 gram of the pure crystallized salt in 1 liter of distilled water. Standardize this solution against the standard potassium permanganate. As the thiosulfate solution does not keep well determine its actual strength at frequent intervals.

3. Starch indicator. Prepare as directed in the section on dissolved oxygen (pp. 65–66).

4. Sodium hydroxide solution. Dissolve 1 part of pure sodium hydroxide in 2 parts of distilled water.

Certain widely practiced deviations from the standard procedure just described are noted in the following paragraphs.

1. Heat the acidified sample to boiling, add the permanganate solution, and digest for two minutes^[16] at boiling temperature. This procedure is facilitated by agitating the liquid constantly with a small current of air to guard against bumping.

2. Same method as No. 1 except that the period of digestion is five minutes.^[121a]

3. Same method as No. 2 except that the permanganate solution is added to the acidified sample when cold, and digestion is continued five minutes after the sample reaches the boiling point. The advantage of this method is that there is included the oxygen-consuming power of the volatile matter present in some sewages and sewage effluents, which is driven off by heat and thus escapes when the test is made in accordance with procedures 1 and 2.

4. Same method as No. 3 except that the period of digestion is 10 minutes.^[63][68c]

5. Digestion of the sample after the acid and permanganate solutions are added is carried out abroad, especially in England, at approximately the room temperature,^{[24a][69a][94f][100a]} apparently to guard against decomposition^[17] of permanganate in the presence of high chloride, for periods of three minutes, fifteen minutes, and four hours; many observers record the oxygen consumed after all three periods, while some record the result only for the four-hour period. At the end of the period of digestion, add 0.5 cc. of potassium iodide solution to discharge the pink color; mix; titrate the liberated iodine with thiosulfate until the yellow color is nearly destroyed, then add a few drops of starch solution and continue titration until the blue color is just discharged. The number of cubic centimeters of potassium permanganate solution in excess of the number of cubic centimeters of sodium thiosulfate solution is equal to parts per million of oxygen consumed.

6. Digestion in alkaline solution^[104] is preferable to digestion in acid solution for brines or waters high in chlorine. Place in a flask 100 cc. of the sample, or if it is of high organic content a smaller portion diluted to 100 cc. Add 0.5 cc. of sodium hydroxide solution and 10 cc. of standard potassium permanganate and digest exactly 30 minutes. Remove the flask, add 5 cc. of sulfuric acid and 10 cc. of the standard ammonium oxalate, and titrate with the standard potassium permanganate as in the acid digestion.

Ignite and weigh a clean platinum dish, and measure into it 100 cc. of the thoroughly shaken sample. Evaporate to dryness on a water bath. Then heat the dish in an oven at 103° C. or 180° C. for one hour. Cool in a desiccator and weigh. The temperature of drying should be mentioned in the report. The increase in weight gives the total solids or residue on evaporation. If 100 cc. of the sample was taken this weight expressed in milligrams and multiplied by 10 is equal to parts per million of residue on evaporation. The residue from waters low in organic matter but relatively high in iron may be used, as a matter of convenience, for the determination of iron.

FIXED RESIDUE AND LOSS ON IGNITION.^[13][96]

The residue from sewages and waters high in organic matter may be ignited to burn off the organic matter, which, with some volatile inorganic matter, constitutes the loss on ignition.

Procedure.—Ignite the residue in the platinum dish at a low red heat. If great accuracy is desired this should be done in an electric muffle furnace or in a radiator, which consists of a platinum or a nickel dish large enough to allow an air space of about half an inch between it and the dish within it, the inner dish being supported by a triangle of platinum wire laid on the bottom of the outer dish. A disc of platinum or nickel foil large enough to cover the outer dish is suspended over the inner dish to radiate the heat into it. The larger dish is heated to bright redness until the residue is white or nearly so. Allow the dish to cool, and moisten the residue with a few drops of distilled water. Dry the residue in the oven, cool in a desiccator, and weigh. The fixed residue on evaporation is the difference between this weight and the weight of the dish.

The loss on ignition is the difference between the total residue on evaporation and the fixed residue on evaporation.

If the odor and color on ignition of some residues give helpful clues to the character of the organic matter record them.

DETERMINATION WITH GOOCH CRUCIBLE.

Reagent.—Prepare a dilute cream of asbestos fibre which has been finely shredded, thoroughly ignited, treated with strong hydrochloric acid for at least 12 hours, and washed with distilled water till free from acid.

Procedure.—1. Prepare a mat of the asbestos fibre 1/16 inch thick in a Gooch crucible. Dry it in an oven at 103 or 180° C., cool and weigh. Filter 1,000 cc. of samples having a turbidity of 50 parts per million or less. If the turbidity is higher use sufficient water to obtain 50 to 100 mg. of suspended matter. Dry for one hour at 103 or 180° C., cool and weigh. Report the temperature at which the residue was dried. If 1,000 cc. is filtered the increase in weight expressed in milligrams is equal to parts per million of suspended matter.

DETERMINATION BY FILTRATION.

The difference between the total solids in filtered and unfiltered portions of a sample may be used as a basis for calculating suspended matter.

DETERMINATION OF VOLUME.

The determination of the volume^[9][69b] of suspended matter in sewages has received considerable attention abroad. Imhoff recommends the use of conical glass vessels holding 1 liter with the lower portions graduated in cubic centimeters. Others recommend centrifuges with sediment tubes.

FIXED RESIDUE AND LOSS ON IGNITION.

Treat the total residue from a filtered sample in the same manner as described for the total residue, and obtain the loss on ignition due to dissolved matter, and by difference the loss on ignition due to suspended matter.

HARDNESS.^[94e]

A water containing certain mineral constituents in solution, chiefly calcium and magnesium, which form insoluble compounds with soap, is said to be hard. Carbon dioxide in water increases the solubility of calcium and magnesium carbonates, forming bicarbonate. If carbon dioxide is removed from the water by boiling the bicarbonate is decomposed and calcium and magnesium are partly precipitated. The proportion of calcium or magnesium carbonate that a water can hold in solution depends on the concentration of carbon dioxide, which in turn depends on the temperature of the water and the proportion of carbon dioxide in the atmosphere with which the water has been in contact. Consequently, when the carbon dioxide is removed from the water by boiling or otherwise the carbonates of calcium and magnesium are partly, but not completely, precipitated, and the hardness of the water is thus diminished and the water is softened to the extent to which these substances are precipitated. The hardness thus removed is called temporary hardness. The hardness which still remains after boiling is due mainly to calcium and magnesium in equilibrium with sulfate, chloride, and nitrate, and residual carbonate, and it is called permanent hardness. Non-carbonate hardness is the hardness caused by sulfates, chlorides, and nitrates of calcium, magnesium, iron, and other metals that form insoluble soaps.

TOTAL HARDNESS BY CALCULATION.

The most accurate method of ascertaining total hardness is to compute it from the results of determinations of calcium and magnesium in the sample. (See methods, pp. 57–58.) Iron and other metals must be included in the calculation if they are present in significant amounts. Total hardness as CaCO₃ equals 2.5 Ca plus 4.1 Mg.

TOTAL HARDNESS BY SOAP METHOD.^[121b]

The determination of hardness by the soap method roughly approximates the amount of calcium and magnesium in a water, though it actually measures the soap-consuming power of the water.

Reagents.—1. Standard calcium chloride solution. Dissolve 0.2 gram of pure calcite (calcium carbonate) in a little dilute hydrochloric acid, being careful to avoid loss of solution by spattering. Evaporate the solution to dryness several times with distilled water to expel excess of acid. Dissolve the residue in distilled water and dilute the solution to 1 liter. One cc. of this dilution is equivalent to 0.2 mg. of calcium carbonate.

2. Standard soap solution. Dissolve 100 grams of dry white Castile soap in 1 liter of 80 per cent alcohol, and allow this solution to stand several days before standardizing. Pure potassium oleate made from lead plaster and potassium carbonate may be used in place of Castile soap.

First method of standardization.—Dilute 20 cc. of the calcium chloride solution in a 250 cc. glass-stoppered bottle to 50 cc. with distilled water which has been recently boiled and cooled. Add soap solution from a burette, 0.2 or 0.3 cc. at a time, shaking the bottle vigorously after each addition until a lather remains unbroken for five minutes over the entire surface of the water while the bottle lies on its side. Then adjust the strength of the stock solution with 70 per cent alcohol so that the resulting diluted soap solution will give a permanent lather when 6.40 cc. of it is properly added to 20 cc. of standard calcium chloride solution diluted to 50 cc. Usually 75 to 100 cc. of the stock soap solution is required to make 1 liter of the standard soap solution. The quantity of calcium carbonate equivalent to each cubic centimeter of the standard soap solution consumed in the titration is indicated in Table 6.

Table 6.—Total hardness in parts per million of CaCO3 for each tenth of a cubic centimeter of soap solution when 50 cc. of the sample is titrated.
Cubic centimeters of soap solution.	0.0.	0.1.	0.2.	0.3.	0.4.	0.5.	0.6.	0.7.	0.8.	0.9.
0.0								0.0	1.6	3.2
1.0	4.8	6.3	7.9	9.5	11.1	12.7	14.3	15.6	16.9	18.2
2.0	19.5	20.8	22.1	23.4	24.7	26.0	27.3	28.6	29.9	31.2
3.0	32.5	33.8	35.1	36.4	37.7	38.0	40.3	41.6	42.9	44.3
4.0	45.7	47.1	48.6	50.0	51.4	52.9	54.3	55.7	57.1	58.6
5.0	60.0	61.4	62.9	64.3	65.7	67.1	68.6	70.0	71.4	72.9
6.0	74.3	75.7	77.1	78.6	80.0	81.4	82.9	84.3	85.7	87.1
7.0	88.6	90.0	91.4	92.9	94.3	95.7	97.1	98.6	100.0	101.5

This table does not provide for the use of so large volume of soap solution for a single determination as former ones because the end-point becomes somewhat obscured in the presence of magnesium if more than 7 cc. is used.

Second method of standardization.—Dilute 100 cc. of the stock soap solution to 1 liter with 70 per cent alcohol. This dilute solution should be of such strength that approximately 6.4 cc. of it will give a permanent lather when 20 cc. of standard calcium chloride solution diluted to 50 cc. with distilled water is titrated with it. Determine the amount of soap solution required to give a permanent lather with 50 cc. of distilled water and with 5, 10, 15, and 20 cc. of standard calcium chloride solution diluted to 50 cc. with distilled water. Finally plot on cross-section paper a curve showing the relation of various quantities of soap solution to corresponding quantities of standard calcium carbonate solution and therefore to parts per million of hardness.

Procedure.—Measure 50 cc. of the water into a 250 cc. bottle and add to it soap solution in small quantities in precisely the same manner as described under the standardization of the soap solution. From the number of cubic centimeters of soap solution used obtain from Table 6 or from the plotted curve the total hardness of the water in parts per million of calcium carbonate.

To avoid mistaking the false or magnesium end-point for the true one^[35] when adding the soap solution to waters containing magnesium salts, read the burette after the titration is apparently finished, and add about 0.5 cc. more of soap solution. If the end-point was due to magnesium the lather will disappear. Soap solution must then be added until the true end-point is reached. Usually the false lather persists for less than five minutes.

If more than 7 cc. of soap solution is required for 50 cc. of the water take less of the sample and dilute it to 50 cc. with distilled water which has been recently boiled and cooled. This step reduces somewhat the disturbing influence of magnesium,^[107a] which consumes more soap than an equivalent weight of calcium.

At best the soap method is not a precise test on account of the different relative amounts of calcium and magnesium in different waters. For hard waters, especially in connection with processes for purification and softening, it is advised that this method be not exclusively used. If the same water is frequently analyzed it may be of assistance to standardize the soap solution against a mixture of calcium and magnesium salts, the relative proportions of which approximate those found in the water.

The strength of the soap solution should be determined from time to time, to make sure that it has not materially changed. Record all results in parts per million of calcium carbonate.

One English degree of hardness, Clark’s scale, is equivalent to 1 grain per Imperial gallon of calcium carbonate. One French degree of hardness is equivalent to 1 part per 100,000 of calcium carbonate. One German degree of hardness is equivalent to 1 part per 100,000 of calcium oxide, and multiplied by 17.9 gives parts per million of calcium carbonate. The relations of these various scales are indicated in Table 7.

Table 7.—Conversion table for hardness.
Unit.	Equivalent.
	Parts per million.	Clark degrees.	French degrees.	German degrees.
One part per million	1.00	0.07	0.10	0.056
One Clark degree	14.3	1.00	1.43	.80
One French degree	10.0	.70	1.00	.56
One German degree	17.9	1.24	1.78	1.00

TOTAL HARDNESS BY SODA REAGENT METHOD.^{[47][74][81][94d]}

Add standard sulfuric acid to 200 cc. of the sample until the alkalinity is neutralized. (See Procedure with methyl orange, p. 37.) Then apply the non-carbonate hardness method (pp. 34–35). This method gives fairly satisfactory estimates of total hardness of hard waters.

TEMPORARY HARDNESS BY TITRATION WITH ACID.

Determine the alkalinity in presence of methyl orange (see p. 37) in the original sample and also in the sample after boiling, cooling, restoring to the original volume with boiled distilled water, and filtering. The difference between the two, if any, is the temporary hardness. This is the most accurate method of determining the temporary hardness of ordinary waters. Iron bicarbonate is included as a part of the temporary hardness.

NON-CARBONATE HARDNESS BY SODA REAGENT METHOD.^{[47][74][81][94d]}

The use of soda reagent does not avoid entirely the error due to solubility of the salts of calcium and magnesium; consequently, if much depends on the results, as in water softening, gravimetric determinations of the calcium and magnesium that remain in solution should be made and a correction should be applied for those amounts.

Reagent.—Prepare soda reagent from equal parts of sodium hydroxide and sodium carbonate. It should be approximately tenth normal.

Procedure.—Measure 200 cc. of the sample and 200 cc. of distilled water into 500 cc. Jena or similar glass Erlenmeyer flasks. Treat the contents of each flask in the following manner. Boil 15 minutes to expel free carbon dioxide. Add 25 cc. of soda reagent. Boil 10 minutes, cool, rinse into 200 cc. graduated flasks, and dilute to 200 cc. with boiled distilled water. Filter, rejecting the first 50 cc., and titrate 50 cc. of each filtrate with N/50 sulfuric acid in the presence of methyl orange or erythrosine indicator. The non-carbonate hardness in parts per million of calcium carbonate is equal to 20 times the difference between the number of cubic centimeters of sulfuric acid required for the soda reagent in distilled water and the number of cubic centimeters of N/50 sulfuric acid required for the soda reagent in the sample.

Water naturally containing bicarbonate and carbonate in excess of calcium and magnesium requires a larger amount of acid to neutralize the sample after it has been treated than is required to neutralize the volume of soda reagent originally added. (See p. 39.)

NON-CARBONATE HARDNESS BY SOAP METHOD.

Non-carbonate hardness may be calculated for waters which are soft or moderately hard in a fairly satisfactory manner by deducting the total alkalinity from the total hardness by the soap method (pp. 31–34). For waters that are very hard, and particularly those that contain much magnesium, this method is not advised.

ALKALINITY.^{[11][18][47][97]}

The alkalinity of a natural water represents its content of carbonate, bicarbonate, borate, silicate, phosphate, and hydroxide. Alkalinity is determined by neutralization with standard sulfuric acid or potassium bisulfate in the presence of phenolphthalein and either methyl orange, erythrosine, or lacmoid as indicators. Methyl orange may be used except in waters containing aluminium sulfate or iron sulfate. The relations between estimates in presence of these indicators and the carbonate, bicarbonate, and hydroxide radicles are indicated in Table 8. The alkalinity of carbonates in the presence of phenolphthalein is different from that in the presence of methyl orange, partly because of loss of carbon dioxide and partly because of defects in phenolphthalein as an indicator in such conditions.

Table 8.—Relations between alkalinity to phenolphthalein and that to methyl orange, erythrosine, or lacmoid, in presence of bicarbonate, carbonate, and hydroxide.
Result of titration.[C]	Value of radicle expressed in terms of calcium carbonate.
	Bicarbonate.	Carbonate.	Hydroxide.
P = 0	T	0	0
P < 1/2T	T - 2P	2P	0
P = 1/2T	0	2P	0
P > 1/2T	0	2(T - P)	2P - T
P = T	0	0	T

Reagents.—1. Sulfuric acid or potassium bisulfate. A N/50 solution.

2. Phenolphthalein indicator. Dissolve 5 grams of a good quality of phenolphthalein in 1 liter of 50 per cent alcohol. Neutralize with N/10 potassium hydroxide. The alcohol should be diluted with boiled distilled water.

3. Methyl orange indicator. Dissolve 0.5 gram of a good grade of methyl orange in 1 liter of distilled water. Keep the solution in the dark.

4. Lacmoid indicator. Dissolve 2.0 grams of lacmoid in 1 liter of 50 per cent alcohol. Dilute the alcohol with freshly boiled distilled water.

5. Erythrosine indicator. Dissolve 0.5 gram of erythrosine (the sodium salt) in 1 liter of freshly boiled distilled water.

PROCEDURE WITH PHENOLPHTHALEIN.

Add 4 drops of phenolphthalein indicator to 50 or 100 cc. of the sample in a white porcelain casserole or an Erlenmeyer flask over a white surface. If the solution becomes colored, hydroxide or normal carbonate is present. Add N/50 sulfuric acid from a burette until the coloration disappears.

The phenolphthalein alkalinity in parts per million of calcium carbonate is equal to the number of cubic centimeters of N/50 sulfuric acid used multiplied by 20 if 50 cc. of the sample was used, or by 10 if 100 cc. was used.

PROCEDURE WITH METHYL ORANGE.

Add 2 drops of methyl orange indicator to 50 or 100 cc. of the sample, or to the solution to which phenolphthalein has been added, in a white porcelain casserole or an Erlenmeyer flask over a white surface. If the solution becomes yellow, hydroxide, normal carbonate, or bicarbonate is present. Add N/50 sulfuric acid from a burette until the faintest pink coloration appears. The methyl orange alkalinity in parts per million of calcium carbonate is equal to the total number of cubic centimeters of N/50 sulfuric acid used multiplied by 20 if 50 cc. of the sample was used, or by 10 if 100 cc. was used.

PROCEDURE WITH LACMOID.

Add 4 drops of lacmoid indicator to 50 or 100 cc. of the sample in a porcelain casserole or an Erlenmeyer flask. Add N/50 sulfuric acid from a burette until within 1 or 2 cc. of the amount necessary for neutralization has been added. Heat the solution until bubbles of steam begin to break at the surface. Remove the dish from the source of heat and continue the titration until a drop of the acid striking the surface of the liquid and sinking to the bottom of the vessel produces no change in the uniform reddish or purple color of the solution. The calculation is the same as for phenolphthalein alkalinity.

PROCEDURE WITH ERYTHROSINE.

Add 5 cc. of neutral chloroform and 1 cc. of erythrosine indicator to 50 or 100 cc. of the sample in a 250 cc. clear glass-stoppered bottle. If the chloroform becomes rose colored on shaking, hydroxide, bicarbonate, or normal carbonate is present. Add N/50 sulfuric acid from a burette until the chloroform becomes colorless. A white surface behind the bottle facilitates detection of a trace of color as the end-point is approached. The calculation is the same as with phenolphthalein alkalinity.

BICARBONATE.

Bicarbonate is present if the alkalinity to phenolphthalein is less than one-half the alkalinity to methyl orange, erythrosine, or lacmoid. The alkalinity to methyl orange, erythrosine, or lacmoid is due entirely to bicarbonate if there is no phenolphthalein alkalinity. If there is phenolphthalein alkalinity the bicarbonate, in terms of calcium carbonate, is equal to the methyl orange, erythrosine, or lacmoid alkalinity minus twice the phenolphthalein alkalinity. Bicarbonate, carbon dioxide as bicarbonate, and half-bound carbon dioxide can be calculated as follows:

Bicarbonate (HCO₃) = 1.22 times the bicarbonate expressed in terms of calcium carbonate.

Carbon dioxide (CO₂) as bicarbonate = 0.88 times the bicarbonate expressed in terms of calcium carbonate.

Half-bound carbon dioxide (CO₂) = 0.44 times the bicarbonate expressed in terms of calcium carbonate.

NORMAL CARBONATE.^[20][94]

Normal carbonate is present if the alkalinity to phenolphthalein is greater than zero but less than the alkalinity to methyl orange, erythrosine, or lacmoid. If the phenolphthalein alkalinity is exactly equal to one-half the methyl orange, erythrosine, or lacmoid alkalinity the alkalinity is due entirely to normal carbonate. If the phenolphthalein alkalinity is less than one-half the methyl orange, erythrosine, or lacmoid alkalinity normal carbonate expressed in terms of calcium carbonate is equal to twice the phenolphthalein alkalinity. If the phenolphthalein alkalinity is greater than one-half the methyl orange, erythrosine, or lacmoid alkalinity the normal carbonate is equal to twice the difference between the methyl orange, erythrosine, or lacmoid alkalinity and the phenolphthalein alkalinity. The carbonate, carbon dioxide as carbonate, and bound carbon dioxide can be calculated as follows:

Carbonate (CO₃) = 0.6 times the normal carbonate expressed in terms of calcium carbonate.

Carbon dioxide as carbonate (CO₂) = 0.44 times the normal carbonate expressed in terms of calcium carbonate.

Bound carbon dioxide (CO₂) is the sum of the carbon dioxide as carbonate and one-half that as bicarbonate.

HYDROXIDE.^[20][94]

If hydroxide, or caustic alkalinity, is present the alkalinity to phenolphthalein is greater than one-half the alkalinity to methyl orange, erythrosine, or lacmoid; the alkalinity is due entirely to hydroxide if the phenolphthalein alkalinity is equal to the methyl orange, erythrosine, or lacmoid alkalinity. If the phenolphthalein alkalinity is more than half and less than all the methyl orange, erythrosine, or lacmoid alkalinity, hydroxide, expressed in terms of calcium carbonate, is equal to twice the phenolphthalein alkalinity minus the methyl orange, erythrosine, or lacmoid alkalinity.

ALKALI CARBONATES.

Waters which contain sodium or potassium carbonates or bicarbonates contain all of their calcium and magnesium as carbonates or bicarbonates. That is, they possess no non-carbonate hardness (sulfates, nitrates or chlorides of calcium and magnesium).

The most accurate method is to determine the total alkalinity by titration with N/50 sulfuric acid, using methyl orange, erythrosine, or lacmoid as an indicator; then determine the calcium and magnesium content; and subtract from the total alkalinity the computed alkalinity due to the calcium and magnesium expressed in terms of calcium carbonate. The remainder is the alkalinity due to carbonates and bicarbonates of sodium and potassium.

This determination may also be made by applying the method, for non-carbonate hardness with soda reagent (see p. 35), and by noting the excess of acid required to neutralize the alkaline carbonates originally present.

With present information as to solubilities of the normal carbonates of calcium and magnesium, it is difficult in their presence to measure slight quantities of carbonates of sodium or potassium.

ACIDITY.^[24d][37]

Waters may have an acid reaction because of the presence of free carbon dioxide, mineral acids, or some of their salts, especially those of iron and aluminium.

Reagents.—1. N/50 sodium carbonate. Dissolve 1.06 grams of anhydrous sodium carbonate in 1 liter of boiled distilled water that has been cooled in an atmosphere free from carbon dioxide. Preserve this solution in bottles of resistant glass protected from the air by tubes filled with soda-lime. One cc. is equivalent to 1 mg. of CaCO₃.

2. N/22 sodium carbonate. Dissolve 2.41 grams of anhydrous sodium carbonate in 1 liter of boiled distilled water that has been cooled in an atmosphere free from carbon dioxide. Preserve this solution in bottles of resistant glass protected from the air by tubes filled with soda-lime. One cc. is equivalent to 1 mg. of CO₂.

3. Phenolphthalein indicator (see p. 36).

4. Methyl orange indicator (see p. 36).

TOTAL ACIDITY.

Procedure.—Add 4 drops of phenolphthalein indicator to 50 or 100 cc. of the sample in a white porcelain casserole or an Erlenmeyer flask over a white surface. Add N/50 sodium carbonate until the solution turns pink. The total acidity in parts per million of calcium carbonate is equal to the number of cubic centimeters of N/50 sodium carbonate used multiplied by 20 if 50 cc. of the sample was used, or by 10 if 100 cc. was used.

FREE CARBON DIOXIDE.^{[20][23][61][87][88][94a][118]}

Carbon dioxide may exist in water in three forms—free carbon dioxide, bicarbonate (pp. 37–38), and carbonate (p. 38). One-half the carbon dioxide as bicarbonate is known as the half-bound carbon dioxide. The carbon dioxide as carbonate plus one-half that as bicarbonate is known as the bound carbon dioxide.

Procedure.—Pour 100 cc. of the sample into a tall narrow vessel, preferably a 100 cc. Nessler tube. Add 10 drops of phenolphthalein indicator, and titrate rapidly with N/22 sodium carbonate, stirring gently, until a faint but permanent pink color is produced. The free carbon dioxide (CO₂) in parts per million is equal to 10 times the number of cubic centimeters of N/22 sodium carbonate used.

Because of the ease with which free carbon dioxide escapes from water, particularly when the gas is present in large amount, a special sample should be collected for this determination, which should preferably be made at the time of collection. If the analysis cannot be made at the time of collection approximate results with water not too high in free carbon dioxide may be obtained on samples collected in bottles completely filled so as to leave no air space under the stopper. Bottled samples should be kept, until tested, at a temperature lower than that of the water when collected. If mineral acids or certain salts are present correction must be made. At best, the results of the titration are uncertain because the proper end-point for correct results differs in color with different types of water.

FREE MINERAL ACIDS.

Procedure.—Add 2 drops of methyl orange indicator to 50 or 100 cc. of the sample in a white porcelain casserole or an Erlenmeyer flask over a white surface. Add N/50 sodium carbonate from a burette until the pink coloration of the solution disappears. The acidity due to free mineral acids, expressed in terms of calcium carbonate, is equal to the number of cubic centimeters of N/50 sodium carbonate used multiplied by 20 if 50 cc. of the sample was used, or by 10 if 100 cc. was used.

MINERAL ACIDS AND SULFATES OF IRON AND ALUMINIUM.^[24d][37]

Procedure.—Modify the method for free mineral acids by titrating the water at boiling temperature in the presence of phenolphthalein indicator. The acidity due to free mineral acids and sulfates of iron and aluminium, expressed in terms of calcium carbonate, is equal to the number of cubic centimeters of N/50 sodium carbonate used multiplied by 20 if 50 cc. of the sample was used, or by 10 if 100 cc. was used.

The acidity due to sulfates of iron and aluminium is equal to the acidity due to mineral acids and sulfates minus the acidity due to mineral acids. The acidity due to ferrous and ferric sulfate can be calculated from the determined amount of these salts (pp. 43–48). The acidity due to aluminium sulfate is equal to the acidity due to total acid sulfates minus that due to iron sulfates.

Acidity shall be reported in parts per million of calcium carbonate (CaCO₃). Sulfate (SO₄) equals parts per million of calcium carbonate multiplied by 0.96.

Carbon dioxide (CO₂) equals parts per million of calcium carbonate multiplied by 0.44.

CHLORIDE.^[16]

Chloride in water and sewage has its origin in common salt, from mineral deposits in the earth, from ocean vapors carried inland by the wind, or from polluting materials like sewage and trade wastes, which contain the salt used in the household and in manufacturing. Comparison of the chloride content of a water with that of other waters in the vicinity known to be unpolluted frequently affords useful information as to its sanitary quality. If, however, the chloride normally exceeds 20 parts per million because of chloride-bearing mineral deposits the chloride content of a water has little sanitary significance.

Reagents.—1. Standard sodium chloride solution. Dissolve 16.48 grams of pure fused sodium chloride in 1 liter of distilled water. Dilute 100 cc. of this stock solution to 1 liter in order to obtain a standard solution each cubic centimeter of which contains 0.001 gram of chloride.

2. Standard silver nitrate solution. Dissolve about 2.40 grams of silver nitrate crystals in 1 liter of distilled water. Standardize this with the standard salt solution, and adjust, correcting for volume (see p. 43), so that 1 cc. will be exactly equivalent to 0.0005 gram of chloride.

3. Potassium chromate indicator. Dissolve 50 grams of neutral potassium chromate in a little distilled water. Add enough silver nitrate to produce a slight red precipitate. Filter and dilute the filtrate to 1 liter with distilled water.

4. Aluminium hydroxide. Electrolyze ammonia-free water, using aluminium electrodes. Wash the precipitate until it is free from chloride, ammonia, and nitrite. Or dissolve 125 grams of potassium or ammonium alum in 1 liter of distilled water. Precipitate the aluminium by adding cautiously ammonium hydroxide. Wash the precipitate in a large jar by successive additions and decantations of distilled water until free from chloride, nitrite, and ammonia.

Procedure.—Add 1 cc. of potassium chromate indicator to 50 cc. of the sample in a 6–inch white porcelain evaporating dish or a 150 cc. Erlenmeyer flask over a white surface. Titrate with the silver nitrate solution under similar conditions of volume, light, and temperature as were used in standardizing the silver nitrate until a faint reddish coloration is perceptible. The detection of the end-point is facilitated by comparison of the contents of the porcelain dish with those of another dish containing the same quantity of potassium chromate indicator in 50 cc. of distilled water. Some analysts prefer to make the titration in a dark-room provided with a yellow light. The end-point is very sharp by electric light and also by daylight with photographic yellow glass. The titration may be made in Nessler tubes^[68a] if the solutions are standardized under similar conditions.

If the amount of chloride is very high use 25 cc., or even a smaller quantity, dilute the volume taken to 50 cc. with distilled water. If the amount of chloride is very low concentrate 250 cc. of the sample to 50 cc. by evaporation. Rotate the liquid to make sure that no residue remains undissolved on the walls of the dish, and, if necessary, use a rubber-tipped glass rod to assist in this operation.

Chloride is determined by some observers by extracting with hot distilled water the residue in the platinum dish in the determination of the residue on evaporation and proceeding as just described. This is permissible if a little sodium carbonate is added before evaporation to prevent loss of chloride through decomposition of magnesium chloride in the residue.

If the sample has a color greater than 30 it should be decolorized by shaking it thoroughly with washed aluminium hydroxide (3 cc. to 500 cc. of the sample) and allowing the precipitate to settle. Make the determination on a portion of the clarified sample, filtered if necessary. If the sample is acid, neutralize it with sodium carbonate; if hydroxide is present, add dilute sulfuric acid until the cold liquid will just discharge the color of phenolphthalein. If the presence of sulfide and sulfocyanate renders it necessary, make proper corrections^[24c][100b] or modifications in treatment.

Make correction for the error due to variations in the volume of the liquid and precipitate by means of the formula^[39] X = 0.003V + 0.02, in which X = the correction in cubic centimeters of silver nitrate solution and V = cubic centimeters of liquid at the end of the titration. If 50 cc. of the sample is titrated chloride (Cl) in parts per million is equal to the number of cubic centimeters of silver nitrate solution multiplied by 10. The correction to be applied is 0.2 cc. unless unusual accuracy is required.

IRON.^[94b][98]

Iron occurs in natural waters in both ferrous and ferric condition, depending on the source of the sample. In ground waters the iron is usually in an unoxidized and soluble condition, sometimes combined with carbonic or sulfuric acid, and also in combination with organic matter. Many waters, especially those that have been exposed to the air, contain the iron in the form of a colloidal hydroxide. Silt-bearing waters often contain much iron in suspension, usually in an oxidized form. Sewages and sewage effluents, particularly those receiving manufacturing wastes, contain various forms of iron of different degrees of solubility, oxidation, and coagulation.

TOTAL IRON.^[59][63b]

COLORIMETRIC METHOD.

Reagents.—1. Standard iron solution. Dissolve 0.7 gram of crystallized ferrous ammonium sulfate in 50 cc. of distilled water to which 20 cc. of dilute sulfuric acid has been added. Warm the solution slightly and add potassium permanganate until the iron is completely oxidized. Dilute the solution to 1 liter. One cc. of the standard solution equals 0.1 mg. Fe.

2. Potassium sulfocyanide solution. Dissolve 20 grams of the salt in 1 liter of distilled water.

3. Dilute hydrochloric acid. One volume of acid (Sp. gr. 1.2) and one volume of distilled water. This shall be free from nitric acid.

4. N/5 potassium permanganate. Dissolve 6.30 grams of the salt in distilled water and dilute to 1 liter.

5. Hydrochloric acid. Concentrated, free from iron.

6. Nitric acid. Concentrated, free from iron.

7. Nitric acid. 5N, free from iron.

First procedure.—Evaporate 100 cc. of the water to dryness, or use the residue left after the determination of residue on evaporation (p. 29). Ignite the residue at a low red heat taking care not to heat it hot enough to make the iron difficultly soluble. Cool the dish and add 5 cc. of concentrated hydrochloric acid. Moisten the inner surface of the dish. Warm the solution for two or three minutes, and again moisten the inner surface of the dish by permitting the hot acid to flow over it. Wash the hot solution from the dish into a 50 cc. Nessler tube, filtering if necessary through paper that has been washed with hot water. Dilute to 50 cc., and add 3 drops of potassium permanganate solution. Add 5 cc. of potassium sulfocyanide solution, mix, and compare with standards.

If it is not convenient to use the residue on evaporation and if the sample is relatively free from organic matter, boil 50 cc. of the sample with 5 cc. of 5N nitric acid for five minutes. Add a few drops of permanganate and 5 cc. of potassium sulfocyanide and compare with standards, using nitric acid in place of hydrochloric acid in the standards. This method is excellent for ground waters. The permanganate and acid liberate chlorine in water high in chloride, and produce a permanent yellow color which interferes with the determination, unless the sample is first diluted to 50 cc. An excess of permanganate, reacting with hydrochloric acid, causes similar trouble. The amounts of hydrochloric acid, 5 cc., and of sulfocyanide, 5 cc., should be approximately measured because more acid lightens the color whereas more sulfocyanide deepens it. This is especially important if permanent standards are used.

Second procedure.—For surface waters containing small amounts of organic matter, the method of Klut^[59] is recommended. Samples containing small amounts of iron should be concentrated, if possible, until at least 0.5 mg. of iron is present in the volume tested. Boil the sample in a beaker with 2 to 3 cc. of concentrated nitric acid free from iron, adding permanganate if necessary to destroy the organic matter. To the hot liquid add ammonia in slight excess and warm until the smell of ammonia is hardly discernible. Filter and wash with water at 70° to 80° C. containing a little ammonia. Dissolve the iron in the beaker and on the filter paper in 5 cc. of concentrated hydrochloric acid, and wash with hot water until the iron is all dissolved, collecting the filtrate in a 50 cc. Nessler tube. Dilute to 50 cc. Add potassium sulfocyanide and determine the iron by comparison with standards.

Comparison with iron standards.

First procedure.—Prepare standards containing amounts of standard iron solution ranging from 0.05 to 4 cc. according to the quantity of iron in the sample. Dilute these amounts with water to about 40 cc. Add 5 cc. of dilute hydrochloric acid and 3 drops of potassium permanganate to each tube and dilute to 50 cc. Add 5 cc. of the potassium sulfocyanide to each of the standard solutions at the same time that it is added to the samples of water under examination, and compare immediately after mixing. If 100 cc. of the sample is used the iron in parts per million is equal to the number of cubic centimeters of the standard iron solution in the standard that the sample matches.

Second procedure.—For a small number of determinations it is more convenient to run the standard iron solution into a Nessler tube containing the acid, distilled water, and potassium sulfocyanide until the color matches that of the sample tested. When determining iron in three or four samples the colors may be matched in the order of their intensity and the volumes of standard iron solution required for each tube may be read from the burette.

Comparison with permanent standards.

Reagents.—1. Platinum solution. Dissolve 2 grams of potassium platinic chloride (PtCl₄.2KCl) in distilled water, add 100 cc. of concentrated hydrochloric acid, and dilute to 1 liter with distilled water.

2. Cobalt solution. Dissolve 24 grams of dry cobaltous chloride crystals (CoCl₂.6H₂O) in a small amount of distilled water, add 100 cc. of strong hydrochloric acid, and dilute to 1 liter with distilled water.

Procedure.—Prepare a series of permanent standards by diluting to 50 cc. with distilled water the amounts of platinum and cobalt solutions, in 50 cc. Nessler tubes, indicated in Table 9. Compare the sample with these standards, and calculate the parts per million of iron.

Table 9.—Preparation of permanent standards for the determination of iron.
Value in standard iron solution.	Platinum solution.	Cobalt solution.
cc.	cc.	cc.
0.0	0	0.0
.1	2	1.0
.3	6	3.0
.5	10	5.0
.7	14	7.5
1.0	20	11.0
1.5	28	17.0
2.0	35	24.0
2.5	39	32.0
3.0	39	43.0
3.5	40	55.0

VOLUMETRIC METHOD.^[24f]

Some samples of sewage and water mixed with trade wastes and mine drainage contain so much iron that it is preferable to use the volumetric method described on page 57 for the determination of both total and dissolved iron, rather than to work with quantities small enough to permit application of the colorimetric methods just described. If iron is present in large quantities in suspension, as in some sewages and septic tank effluents, it may be filtered off and the residue washed, ignited, and fused with potassium and sodium carbonate. The fusion is then extracted with hydrochloric acid and the iron determined as on page 57.

Samples containing much organic matter should be evaporated to dryness with 0.5 cc. of concentrated sulfuric acid and the residue then ignited before estimation of iron.

DISSOLVED IRON.

Determine, by the method described for total iron, the iron in the sample after filtration. Iron may precipitate from some samples during filtration.

SUSPENDED IRON.

The suspended iron is the difference between total iron in the unfiltered sample and dissolved iron in the filtered sample.

FERROUS IRON.^[24e]

Determine the total ferrous iron in an unfiltered sample and the dissolved ferrous iron in a filtered sample.

Reagents.—1. Standard iron solution. Dissolve 0.7 gram of crystallized ferrous ammonium sulfate in a large volume of freshly boiled distilled water to which 10 cc. of dilute sulfuric acid has been added and dilute to 1 liter. This solution should be freshly prepared when needed. One cc. of this standard solution contains 0.1 mg. of Fe.

2. Potassium ferricyanide solution. Dissolve 5 grams of the salt in 1 liter of distilled water. Use a freshly prepared solution.

3. Dilute sulfuric acid. Dilute 1 part of sulfuric acid, specific gravity 1.84, with 5 parts of distilled water.

Procedure.—Add 10 cc. of dilute sulfuric acid to 50 cc. of the sample, remove the suspended matter by filtration if necessary, and add 15 cc. of potassium ferricyanide solution. Dilute the solution to 100 cc. with distilled water. Compare the color developed in the sample with that in standards made at the same time from the ferrous iron solution. Place in 100 cc. Nessler tubes, in the following order, 75 cc. of distilled water, 10 cc. of dilute sulfuric acid, and 15 cc. of potassium ferricyanide solution, and mix well the contents of each tube. Prepare as many tubes in this way as are needed. Add various quantities of standard ferrous iron solution to several tubes, mix well, and compare the resulting colors with the samples immediately.

FERRIC IRON.

The amount of ferric iron in solution and suspension is equal to the difference between the total iron and the ferrous iron obtained by the methods described.

MANGANESE.

If the sample contains less than 10 parts per million of manganese, use a colorimetric method in which the manganous salt is oxidized to permanganate and the color produced thereby is compared with that of a standard solution similarly treated. The persulfate method and the bismuthate method are suitable. If the sample contains more than 10 parts per million of manganese it is sometimes preferable to use a volumetric or gravimetric method.

PERSULFATE METHOD.

Reagents.—1. Nitric acid. Dilute concentrated nitric acid with an equal volume of distilled water. Free the diluted acid from brown oxides of nitrogen by aeration.

2. Silver nitrate. Dissolve 20 grams of silver nitrate in 1 liter of distilled water.

3. Standard manganous sulfate. Dissolve 0.288 gram of purest potassium permanganate in about 100 cc. of distilled water. Acidify the solution with sulfuric acid and heat to boiling. Add slowly a sufficient quantity of dilute solution of oxalic acid to discharge the color. Cool and dilute to 1 liter. One cc. of this solution contains 0.1 mg. of manganese.

4. Ammonium persulfate. Crystals, free from chloride.

Procedure.—Use an amount of the sample that contains not more than 0.2 mg. of manganese. Add 2 cc. of nitric acid and boil down to about 50 cc. Precipitate the chloride with silver nitrate solution, adding at least 1 cc. in excess. Shake and heat to coagulate the precipitate, and filter. A sample that contains much chloride should be evaporated with a few drops of sulfuric acid until white fumes appear and then diluted before the nitric acid and silver nitrate are added as directed above. If the sample is highly colored by organic matter it should be evaporated with sulfuric acid, and the residue ignited and dissolved in dilute nitric acid. Add about 0.5 gram of ammonium persulfate crystals and warm the solution until the maximum permanganate color is developed. This usually takes about ten minutes. At the same time prepare standards by diluting portions of 0.2, 0.4, 0.6 cc., etc. of the standard manganous sulfate solution to about 50 cc. and treating them exactly as the sample was treated. Transfer the sample and the standards to 50 cc. Nessler tubes, and compare the colors immediately. Manganese in parts per million is equal to the number of cubic centimeters of standard manganous sulfate solution in the tube that the sample matches multiplied by 100, divided by the number of cubic centimeters of the sample used.

BISMUTHATE METHOD.^[2a][113]

Reagents.—1. Nitric acid. Dilute 1 part of concentrated nitric acid with 4 parts of distilled water. Free the dilute acid from brown oxides of nitrogen by aeration.

2. Sulfuric acid. Dilute 1 part of concentrated sulfuric acid with 3 parts of distilled water.

3. Dilute sulfuric acid. Dilute 25 cc. of concentrated acid to 1 liter with distilled water. Add enough permanganate solution to color faintly the dilute acid.

4. Standard manganous sulfate. The standard solution of manganous sulfate prepared as described under persulfate method (p. 48) should be used and the standards should be prepared by following the same procedure as is used for the sample. This solution is more permanent than a solution of potassium permanganate, which may, however, be used. To prepare it dissolve 0.288 gram of potassium permanganate in distilled water and dilute the solution to 1 liter.

5. Sodium bismuthate. Purest dry salt.

Procedure.—Use an amount of the sample that contains not more than 0.2 mg. of manganese. Add 0.5 cc. of sulfuric acid and evaporate to dryness. Heat until the sulfuric acid is volatilized and ignite the residue. Dissolve in 40 cc. of nitric acid, add about 0.5 gram of sodium bismuthate, and heat until the permanganate color disappears. Add a few drops of a solution of ammonium or sodium bisulfate to clear the solution and again boil to expel oxides of nitrogen. Remove from the source of heat, cool to 20° C., again add 0.5 gram of sodium bismuthate, and stir. When the maximum permanganate color has developed, filter through an alundum or Gooch crucible containing an asbestos mat ignited and washed with potassium permanganate. Wash the precipitate with dilute sulfuric acid until the washings are colorless. Transfer the filtrate to a 50 cc. Nessler tube and compare the color of it with that of standards prepared from the potassium permanganate solution. To prepare the standards, dilute portions of 0.2, 0.4, 0.6 cc., etc. of the permanganate solution to 50 cc. with dilute sulfuric acid. The content of manganese is calculated as described under persulfate method (p. 49).

LEAD, ZINC, COPPER, AND TIN.^[7][60]

Determinations of lead, zinc, copper, and tin are important in certain mining regions and in places where the water has a solvent action on pipes and other containers. The use of certain “germicides” also makes it necessary to test for some of these metals.

Lead, zinc, and copper may be determined colorimetrically or electrolytically. The colorimetric methods are not so accurate as a combination of both, and are chiefly of value as qualitative tests.

It is possible to make a rough estimation of the amount of lead in clear waters by acidifying with acetic acid, saturating with hydrogen sulfide, and comparing the color produced with that produced by standard lead solutions in Nessler tubes, treated in similar manner. This method, however, is not applicable if the water is colored or contains iron.

Reagents.—1. Standard lead solution. Dissolve 1.60 grams of lead nitrate (Pb(NO₃)₂) in 1 liter of distilled water. One cc. of this solution contains 1 mg. of lead (Pb). As a check it is desirable to determine lead as sulfate in a measured portion of this solution.

2. Standard copper solution. Dissolve about 0.8 gram of copper sulfate crystals (CuSO₄.5H₂O) in water and, after the addition of 1 cc. of concentrated sulfuric acid, dilute the solution to 1 liter. Determine the copper in 100 cc. of this solution in the usual way by electrolytic deposition. Dilute the solution so that 1 cc. contains 0.2 milligram copper (Cu). This solution is permanent.

3. Ammonium chloride. Twenty-five per cent solution.

4. Ammonium acetate. Fifty per cent solution.

5. Ammonium hydroxide. (Sp. gr. 0.96.)

6. Hydrogen sulfide. Saturated solution.

7. Potassium sulfide. An alkaline solution of potassium sulfide made by mixing equal volumes of 10 per cent potassium hydroxide and a saturated aqueous solution of hydrogen sulfide.

8. Potassium oxalate. Crystals.

9. Potassium sulfate. Crystals.

10. Alcohol. Ninety-five per cent.

11. Alcohol. Fifty per cent.

12. Acetic acid. Fifty per cent.

13. Nitric acid. Concentrated acid (Sp. gr. 1.42).

14. Nitric acid. Dilute 1 part of the concentrated acid to 10 parts with distilled water.

15. Hydrochloric acid. (Sp. gr. 1.20.)

16. Sulfuric acid. Concentrated acid (Sp. gr. 1.84).

17. Sulfuric acid. Dilute the concentrated acid with an equal volume of distilled water.

18. Urea. Crystals.

LEAD.

Concentrate (1)^[D] rapidly by boiling in a 7–inch porcelain dish over a free flame 3 or 4 liters of the sample to be tested, or more if very small amounts of the metals are present, to a volume of about 30 cc. Add 10 or 15 cc. of ammonium chloride solution to assist in the separation of the sulfides, then add a few drops of concentrated ammonium hydroxide, and saturate with hydrogen sulfide. Allow to stand some time, preferably over night, add a little more ammonium hydroxide and hydrogen sulfide, boil the contents of the dish a few minutes, and filter. The precipitate (2) may consist of lead, zinc, copper, and iron sulfides and the suspended organic matter. The soluble coloring matter is in the filtrate (3). Wash the precipitate a few times with hot water, place the precipitate and the filter paper in the original dish and boil with dilute nitric acid, rubbing down the sides of the dish, if necessary, to detach any adhering sulfide precipitate. After again filtering and washing several times with hot water, evaporate the filtrate and washings in the original dish to a bulk of 10 to 15 cc., cool, add 5 cc. of concentrated sulfuric acid, and heat until copious fumes of sulfuric acid are evolved.

D. The numbers in parentheses refer to tables 10–12, pages 55–56.