Cooley's Cyclopaedia of Practical Receipts Volume Ii Part 294

Ferrocyanide of pota.s.sium gives a dark blue precipitate in water containing a ferric salt; and a white one, turning blue by exposure to the air, in water containing a ferrous salt. 12. If sulphuric acid be run into water and allowed to cool, and a crystal of sulphate of iron dropped into the water, a dark brown cloud round the crystal indicates nitrates; the bleaching of indigo added to the hot mixture of equal parts water and pure oil of vitriol also indicates the presence of these salts. 13. Sulphuric acid or sulphates is indicated by a soluble salt of barium throwing down a white precipitate insoluble in nitric acid.

_Water, Quant.i.tative a.n.a.lysis of._--The quant.i.tative a.n.a.lysis of potable water is confined to the following: total residue, hardness temporary and permanent, chlorine, ammonia, nitrates and nitrites, and organic matter.

Of these, all but the first two are intended to throw light on the organic contamination of the water. Chlorine, ammonia, and nitrates and nitrites are in themselves innocuous substances, but are estimated because they supplement the somewhat imperfect information obtained from the organic matter itself. A sewage-polluted supply being an agent in propagating zymotic diseases, a knowledge of the source of the organic matter in a water is of the highest importance.

Before pa.s.sing to the mode of estimating the above items it may be desirable to explain the object of each a.n.a.lysis and the interpretation which may be placed on the results.

_Total solid residue_ includes all the substance, organic or mineral, dissolved in the water. Everything beyond the two gases which enter into the combination of the water being useless, the 'residue' of a water is sometimes called the 'total solid impurity.' The less residue left by a water on evaporation the better, but a water need not be objected to for drinking purposes till the residue reaches 40 grains per gallon. For raising steam a water should not contain more than 20 grains, and should be, if possible, much less.

_The hardness_, or soap-wasting power of a water, is chiefly determined on economic grounds. Unless the hardness is very excessive, the hardness or softness of the water does not appear to materially affect the health of the consumer. Hardness is caused by salts of lime and magnesia. If the property of hardness be caused by the presence of bicarbonates of the above substances, the water is said to be 'temporarily' hard, for by boiling or adding lime as above described, the hardness may be reduced without affecting the potability of the supply; but when the hardness is due to calcium or magnesium sulphates it is called 'permanent' hardness, for it is not then practicable to remove the hardening ingredients without adding some more objectionable substance. The average hardness of the four cla.s.ses of pure water is shown in the a.n.a.lyses given above. Thames water has a total hardness of 15, Loch Katrine water, as supplied to Glasgow, 070, on Clarke's scale.

_Chlorine._--Except in places near the sea, or in salt-bearing strata, an unpolluted water does not contain more than the merest trace of chlorine.

Sewage, however, contains a large quant.i.ty of chlorine as sodic chloride (common salt) derived from the salt used in cooking, &c. Hence a mixture of sewage with water becomes known by the quant.i.ty of chlorine present. It is not safe to drink a water containing such an excessive quant.i.ty of chlorine as 4 grains per gallon. The chlorine in Ullswater and Thames water is 7 and 11 grains per gallon respectively. Sewage has about 8 grains on the average.

_Ammonia._--This determination acquires significance because it is one of the early substances produced by the decomposition of animal matter. It therefore indicates, when present in large quant.i.ties, _recent_ contamination by sewage. Rain always contains a small amount of ammonia, and deep wells occasionally show ammonia derived from the reduction of nitrates by the oxygen-seeking organic matter. The above inferences must, therefore, be applied with caution.

_Nitrates and Nitrites_ result from the oxidation of animal matter.

Vegetable substances, under like conditions, yield none or but mere traces of these compounds. The presence of nitrates is a most unfavorable sign in a shallow well or river water, because the conditions to which these waters are subjected are so variable that there is a constant liability of the purifying processes diminishing, and allowing the sewage, now only represented by innoxious nitrates, to appear in its dangerous, unoxidised condition.

Dr Frankland takes the sum of the nitrogen existing in the water as ammonia and as nitrites and nitrates, as a sort of measure of the minimum amount of animal or sewage matter destroyed. The amount due to sewage or animal matter is considered to be all over 032 part per 100,000 (or 022 gr. per gallon), which is the average of 'inorganic nitrogen' natural to unpolluted rain water. Dr Frankland also expresses this 'previous sewage or animal contamination,' in terms of London sewage containing 10 parts of nitrogen in 100,000 parts of liquid, by multiplying the above-named corrected sum by 10,000. Thus, a water containing 1 part per 100,000 (7 gr. per gall.) of 'inorganic nitrogen' would have a 'previous sewage or animal contamination' of 9680 parts per 100,000, for it would have required 100,000 {(1 - 032)/10} = 9680 parts of London sewage to produce an amount of nitrogen equal to that found by a.n.a.lysis. A water which contains over 20,000 parts of previous sewage contamination (15 grains of inorganic nitrogen) is said to be dangerous. All other waters containing more inorganic nitrogen than in rain are said to be 'doubtful'

except springs and deep well waters containing less than 10,000 parts of previous sewage contamination per 100,000, and such shallow wells and running water which from their source may be taken to be free from sewage.

_Organic matter._--There is no method by which the actual weight of organic matter can be determined, still less is it possible to say how much is likely to be actually injurious organic matter, but there are several means of measuring the proportionate amount of organic contamination.

Dr Frankland determines the amount of carbon and nitrogen in the organic matter. The smaller the amount of these elements the better the water, and the less the amount of nitrogen, especially in proportion to organic carbon, the less chance of _animal_ matter. A good drinking water will not have more than 2 parts in 100,000 (14 gr. per gall.) of carbon, or 03 part of organic nitrogen in 100,000 parts (02 gr. per gall.) of the water. The amount of putrescent matter may be estimated by the amount of oxygen consumed in destroying it. Dr Tidy ('Chem. Soc. Jour.,' January, 1879) considers that, speaking generally, waters requiring 05 part per 100,000 (035 gr. per gall.) to be of great organic purity; 15 part (1 gr. per gall.) waters of medium purity; waters of doubtful purity, from 15 to 21 part per 100,000 (15 gr. per gallon). Impure waters, all above 15 gr. per gall.

The proportion of alb.u.minous substances present is measured by Mr w.a.n.klyn by the amount of ammonia set free by alkaline permanganate. A water containing over 15 part per million alb.u.minoid ammonia condemns a water absolutely ('w.a.n.klyn's Water a.n.a.lysis,' 4th edit., p. 54); 10 part per million with little free ammonia, or 05 part alb.u.minoid ammonia with much free ammonia, is 'suspicious.' A water with less than 05 part alb.u.minoid ammonia belongs to the cla.s.s of very pure waters.

Of course the above data are not hard and fast lines, but serve as aid to a judgment which may be modified by other circ.u.mstances connected with the a.n.a.lysis, and the source of the water.

_Methods of a.n.a.lysis. Total solid residue._--1000 grains are evaporated to dryness in a platinum dish over a water bath and residue dried in an oven at 212 F. for an hour, or until the weight is constant. The increase in weight of the platinum vessel multiplied by 70 gives the number of grains of total solid residue per gallon.

_Hardness_ is determined by a solution of soap of which 320 grain-measure will soften a water of 16 of hardness. Each degree of hardness represents an amount of soap-destroying matter equivalent to 1 grain of chalk per gallon. 1000 measured grains of the water are measured into a narrow-mouthed six or eight ounce stoppered bottle, then well shaken, and the air sucked out by means of a piece of gla.s.s tube. The standard soap solution is now run in 10 grains at a time, shaking well between each addition until there is formed over the whole surface a lather which, when the bottle is placed upon its side, shall last just five minutes. The number of grain-measures used will indicate the hardness of the water by reference to Table A. Should, however, the permanent lather not be formed before 320 measures of soap solution have been added, a second trial must be made, in which only 500 grain-measures of the water are taken, to which a like amount of recently-boiled distilled water is added. The degree of hardness now obtained must be multiplied by 2. With very hard waters it is necessary to dilute still further, say 250 grains to 750 of distilled, and multiplying the result by 4. If the number of soap-measures does not correspond with any degree on the table, observe which numbers it falls between. The degree corresponding to the lower of these soap volumes will be the whole number in the answer; the fraction will be the difference between the observed number of measures and the next lower on the table, divided by the difference (given in column 3) between the figure above and below it. Thus, if 14 measures were used the hardness would be 62, 136 measures being equivalent to 6 degrees, and the fraction being {14 - 136}/{136 - 116} = 4/20 = 2.

The hardness of the water in the natural state is the 'total hardness.' By boiling for an hour and making up loss by evaporation with boiled distilled water and again determining the hardness, the 'permanent hardness' is found. That which has been removed by the boiling is the temporary hardness.

TABLE A.

Soap test measures corresponding to one thousand measures of water of each degree of hardness.

Degree of Soap test Difference.

hardness. measures.

0 14 18 1 32 22 2 54 22 3 76 20 4 96 20 5 116 20 6 136 20 7 156 19 8 175 19 9 194 19 10 213 18 11 231 18 12 249 18 13 267 18 14 285 18 15 303 17 16 320 --

The standard water of 16 of hardness is thus made:--Pure carbonate of calcium (Iceland spar) is weighed out into a porcelain or platinum dish in the proportion of 16 grains for a gallon of solution. It is dissolved in weak hydrochloric acid, and the whole cautiously evaporated to dryness over a water bath, then re-dissolved in water and again evaporated to drive off any excess of acid. The dish is covered with a gla.s.s during the operation to prevent loss by spirting. The resulting neutral chloride of calcium is dissolved in a gallon of pure distilled water if 16 grains were weighed out, or a proportionate quant.i.ty in other cases. The soap solution can be made by dissolving good curd soap in weak methylated spirit in the proportion of one ounce of soap to the gallon. A potash soap made as follows is, however, less liable to change: 150 grains of lead plaster (Emplastrum plumbi, B. P.) and 40 grains of dry pota.s.sic carbonate are rubbed together in a mortar and repeatedly extracted with small portions of methylated spirit, triturating the ma.s.s meanwhile, till about a pint of spirit has been used; filter and add an equal bulk of recently boiled distilled water. Whichever method is followed the clear solution has now to be standardised by the 'water of 16 of hardness.' 1000 grains of the water of 16 of hardness are placed into a bottle, and this soap solution is run in from a burette until a permanent lather is formed. The soap solution must be fortified by strong soap solution or diluted with alcohol till 320 measures produce a lather permanent for five minutes in 1000 grain-measures of water of 16 of hardness.

_Chlorine._ To 1000 grains of the water add a drop or two of neutral chromate of sodium, so as to tinge the water yellow; run in standard nitrate of silver till the liquid acquires a very faint red tinge, showing that all the chlorine has been precipitated and that red silver chromate is beginning to be formed. The number of grains of standard solution divided by 100 will give the grains of chlorine in one gallon of the water.

The standard solution is prepared by dissolving pure nitrate of silver in the proportion of 4790 grains to one gallon of distilled water.

_Ammonia_ is always carried out as described in the account of Messrs w.a.n.klyn and Chapman's process.

_Nitrate and Nitrites._--These substances can be most expeditiously estimated by the indigo process as follows: 200 grain-measures of the water are placed in a flask and a little of a standard solution of indigo added thereto; twice the volume of pure sulphuric acid is then suddenly poured in from a measuring cylinder, and the whole shaken. The temperature rises immediately to about 270 Fahr., and the blue colour will probably be immediately discharged; more indigo, therefore, must be rapidly run in till a brown-green tint shows itself. This gives the trial estimation, but the maximum amount of indigo is only used up when all the indigo is added previous to the addition of acid; hence a second experiment is now started, and an amount equal to that previously used run in at once, and on it is poured exactly twice as much sulphuric acid as there is water and indigo in solution. The second result will be somewhat higher than the first. If the solutions below mentioned be used, the amount of indigo required by the 200 grains of water divided by the number of grains of indigo required to bleach 200 c.c. of standard nitre represents the grains per gallon of nitrogen as nitrates and nitrites. The standardising of the indigo with the nitrate solution is performed exactly as for an actual water. The requisites are a solution of pure pota.s.sium nitrate of known strength, say 14442 gr. of nitre (equivalent to 2 gr. of nitrogen or 9 gr. of nitric acid) in a gallon of distilled water. 2. A solution of indigo made by dissolving soluble indigo carmine in distilled water in such a proportion that 200 gr. is about equal to 200 gr. of nitre solution. 3. Strong pure oil of vitriol; it must be free from nitrous compounds, not become turbid when diluted, and its specific gravity not be less than 184. It is important to maintain the same proportion of acid, and not to allow the temperature to fall below 250 F. throughout the experiment.

Messrs w.a.n.klyn and Chapman's aluminium method is also a very convenient process. 2000 grains of the water are placed in a retort and half as much of a solution of 10 per cent. soda added. The soda solution is made from sodium soda and the absence of nitrates is secured by boiling the liquid with a piece of aluminium. Half the contents of the retort are distilled over and the residue cooled. A piece of aluminium foil of about six square inches area is tied to a piece of clean gla.s.s rod and sunk in the liquid.

The neck of the retort is guarded by a tube containing fragments of gla.s.s moistened with hydrochloric acid; it is sloped, so that any liquid spurted into the neck will flow back into the retort. After resting several hours the neck of the retort is washed down with pure water, the contents of the tube are transferred to the retort, and the contents distilled over, down to about an ounce in two or three ounces water placed as a receiver. The contents of the receiver are made up to 200 grains and the ammonia is estimated in one half by Nessler's test as below described.

An exceedingly accurate eudiometric method has also been devised by Dr Frankland, based on Crum's observations, that a highly concentrated solution of nitrates, when vigorously agitated with mercury and an excess of concentrated pure sulphuric acid, yields all its nitrogen from the nitrates and nitrites, as nitric oxide, a compound occupying twice the volume of the nitrogen as nitrates. The weight of gas is easily calculated from the volume measured ('Journal Chem. Soc.,' March, 1868).

_Organic Contamination; means of estimating._--Messrs w.a.n.klyn and Chapman's method is most generally employed. It depends on the conversion of the nitrogen of the organic matter into ammonia and the employment of Nessler's test to estimate this ammonia.

_Nessler's Test._ 500 gr. of iodide of pota.s.sium are dissolved in a small quant.i.ty of hot distilled water, and to this is gradually added a cold saturated solution of mercuric chloride till the precipitate produced ceases to be dissolved upon stirring. To render this alkaline, add 2000 gr. of pota.s.sic hydrate and dilute the volume to 10,000 grain measures. A little more saturated mercuric chloride is added, and the whole allowed to settle, and the clear liquid decanted off. The test should have a slightly yellowish tint. If colourless, it is not sensitive, and more mercuric chloride must be added.

_Standard Ammonia Solution._--Dissolve 27164 gr. of pure sulphate of ammonium in 1 gall. of distilled water. For use dilute 100 gr. to 1000 gr.

It will then contain 1 gr. of ammonia in 100,000 of water.

In order to estimate ammonia several six-ounce tall gla.s.s cylinders, free from colour, are graduated at 1000 grains. One of these is filled up to the graduation mark with the ammonia to be estimated, and about 30 gr. of Nessler's reagent added from a pipette. The coloration produced is noted, a second cylinder is filled nearly to the mark with distilled water, and what is thought sufficient ammonia to produce a similar colour to the first run in, and the whole made up to 1000 gr., and 30 gr. of Nessler added; if after standing five minutes the colour in the second is the same as in the water examined, the quant.i.ty of ammonia they contain will be equal; but if this is not the case a second trial must be made, using more or less standard ammonia as the intensity of colour is greater or less than the first. After a little experience, more than two trials are rarely necessary.

_Examination._ (_a_) _Free Ammonia._ 7000 grains (a deci-gallon) of the water to be a.n.a.lysed is placed in a tabulated retort, and to it is added half an ounce of a supersaturated solution of carbonate of soda, made by dissolving ignited carbonate of soda in water free from ammonia. The contents are distilled over in two portions of 1000 grains each, and the second Nesslerised; if it contains no ammonia, the distillation may be stopped; if it does, the distillation must be continued and tested in portions of 500 grains till the ammonia no longer can be detected. If there is much ammonia in the cylinder of the second 1000 grains the first will probably contain too much to be conveniently estimated, and therefore an aliquot part diluted to 1000 grains with distilled water free from ammonia should be used. The sum of the ammonia in these different portions multiplied by 10 gives grains per gallon.

(_b_) _Alb.u.minoid Ammonia._ To the retort, after all the free ammonia has been driven off, one ounce of a solution of hydrate and permanganate of pota.s.sium of a strength of 2000 gr. of hydrate of pota.s.sium and 80 gr. of permanganate to 10,000 gr. of water is added, and the distillation continued until no more ammonia comes over, collecting the distillate in portions of 1000 c.c. as before. The sum is the alb.u.minoid ammonia derived from the nitrogenous organic matter.

It is of course essential that the utmost care be taken to remove by rinsing or distillation all traces of ammonia from apparatus employed.

Water which has been distilled till free from ammonia should alone be used in estimations and preparations of solutions, and the alkaline permanganate should lie boiled for a short time when made to expel ammonia.

_"Oxygen" Process._ This is a useful process when comparing waters of similar origin. It is probably a more reliable measure of the _putrescent_ matter present than the _total_ organic contamination. It is essential that the oxidizing agent pota.s.sium permanganate be added in excess and allowed to stand three hours. The following method is very delicate (_vide_ Dr Tidy on Potable Waters, 'Chem. Soc. Journ.,' January, 1879.)

Cleanse with sulphuric acid and with tap water two flasks and place in one 500 septems (1/20 gall.) of the water, in the other an equal quant.i.ty of distilled water. Add to each 20 septems (140 gr.) of sulphuric acid (1 part pure acid to 3 of distilled water) and 20 septems of pota.s.sium permanganate and allow to rest for 3 hours. Then add to each flask a couple of drops of an aqueous solution of pota.s.sic iodide (1 in 10) when iodine is liberated equivalent to the amount of permanganate unacted on by the waters. Observe the amount of a sodic hyposulphite solution (54 gr.

in 7000 gr.) which must be added to each to remove this free iodine (judging of the exact spot by adding towards the end of the experiment a few drops of starch).

The strength of the pota.s.sic permanganate solution is 2 gr. of the salt in 7000 gr. to 1/10 gall.; therefore the 20 septems will contain 04 gr.

permanganate, equivalent to 01 of available oxygen. The experiment (A) with the amount of hyposulphite used up for the blank distilled water shows the amount of hyposulphite equivalent to 20 septems or 01 gr. of oxygen. Therefore the amount of oxygen unconsumed in the water (B) to be examined was (B/A) 01 and the amount (C) actually used up was (A - B)/A 01 for 500 septems (1/20th gall). Then the oxygen consumed per gallon would be A-B x 2 / A. It is necessary to perform this standardising of hyposulphite with every series of experiment on account of its tendency to change. Dr Tidy recommends that in addition to the three hours' experiment one of a single hour duration be executed. The higher the proportion of oxygen consumed in one hour to the oxygen consumed in three hours the worse the water.

Nitrites, sulphuretted hydrogen, and ferrous salts interfere with this test, and there appears to be a different ratio between the oxygen consumed and the amount of organic matter according to the amount of oxidation that has already taken place. The organic matter of deep wells is proportionately least acted upon.

_Combustion methods._--The "Frankland and Armstrong process" consists in burning with oxide of copper in vacuo the residue left on evaporating the water, and collecting and measuring in a suitable gas apparatus the carbonic acid, and nitrogen, and nitric oxide proceeding from the organic matter. From these estimations are calculated the organic carbon and nitrogen.

This method, though forming the most accurate means of measuring organic contamination, is not in general use in consequence of the difficulties attending Dr Frankland's method of a.n.a.lysis. Professor Dittmar and Drs Dupre and Hake have lately introduced processes by which the same results may be obtained without necessitating the use of expensive gas apparatus.

_Dittmar's Carbon._--Concentrate a suitable quant.i.ty (say 10,000 gr.) in a pear-shaped flask, and, after adding some saturated solution of sulphurous acid to expel carbonates and nitrates, evaporate to dryness in a gla.s.s dish on a water bath. Transfer the residue from the dish to a porcelain or platinum boat, and introduce it into the tail end of a combustion tube, filled three fourths of its length with oxide of copper, and having a roll of silver gauze in the front part of the tube. Previous to the boat being put in, this tube is heated to redness, and a stream of air, freed from carbonic acid, pa.s.sed through it till the gas which comes out no longer renders clear baryta water turbid. The combustion tube has attached in front a small V-Shaped tube charged with chromic acid, dissolved in 60 per cent. sulphuric acid. To it is permanently fixed a small tube filled with calcic chloride, and in front of all is a small-weighed U-tube the first three fourths of which is filled with soda lime, and the other fourth with calcic chloride. On turning the gas on gradually from the front to the tail the residue is at last reached, and burnt in the stream of pure air.

The carbonic acid given off, after being freed from sulphurous anhydride by pa.s.sing through the chromic acid solution and of moisture by the calcic chloride, pa.s.ses into the soda lime tube and is absorbed. The increase in weight multiplied by 3/11 gives the amount of carbon in the amount of water taken.

_Dittmar's Nitrogen._ An amount of water, about half that taken for the carbon, is evaporated in a similar way. The residue is transferred to a large copper or silver boat, and mixed with about 50 grains of soda made from pure sodium, or with a mixture of soda and baryta, and burnt in a stream of hydrogen in a short combustion tube, which is closed in front by a nitrogen absorption bulb charged with exceedingly weak acidulated water.

The amount of ammonia given off is estimated by the Nessler test as described under "Ammonia." Subtracting the amount of inorganic ammonia the residue multiplied by 14/17 yields the quant.i.ty of organic nitrogen in that volume of water.

A few blank experiments must be made to observe and allow correction for the amount of experimental error.

_Carbon method of Drs Dupre and Hake._[261]--This method appears to be very accurate, but it necessitates a number of minute precautions, which cannot here be particularised. A residue is obtained by evaporating the water either in the ordinary hemispherical gla.s.s dish, or in an exceedingly thin silver one, which after being ignited is supported in a platinum hemisphere of convenient size. At the close of the evaporation this dish is crumpled up without being handled and introduced into a combustion tube, similar to that described under Dittmar's process. The carbonic acid is absorbed in bright baric hydrate solution, and the precipitated baric carbonate is, with suitable precautions to prevent access of impure air, collected on a filter and washed. It is dried and weighed. The result divided by 194 gives the weight of organic carbon. As another method of estimating the carbon the authors propose to compare the turbidity produced by the carbonic acid evolved from the combustion of the residue in solutions of basic acetate of lead with that produced by known quant.i.ty of carbonic acid.