摘要:

77 XI.-On t?Le Analysis of Potable Waters. By E. FRANKLAND Esq. F.R.X. and H. E. ARMSTRONG ALTHOUGH the analysis of potable waters has received no incon-siderable amount of attention the subject is surrounded with such formidable difficulties that at the time we undertook its inveetigation it was generally regarded as one of the least satisfactory of analytical operations. The diEculties concen- trate theinselves chiefly upon the determination of the organic niatters contained in all potable natural waters and of the mineral compounds derived from the decomposition of these organic matters viz. nitrous and nitric acids and ammonia; the processes used for the estimation of the remaining mineral ingredients leaving little to be desired. In the year 1856 Hofmann and Blyth" drew attention to the defects of' the processes then in use and showed that the loss experienced on ignition by the solid residue of a water could be made more nearly than before to represent the organic matters by excluding from this loss ammoniacal salts moisture and hydrochloric acid an improvement which they effected by the addition of a known weight of sodic carbonate to the water before evaporation.These chemists also pointed out the great desirability of determining the amount of iiitrogen entering into the composition of the organic matters contained in waters although they did not succeed in devising any process by which this could be accomplished. In 1864 W eltzient described a new process for determining the amount of nitric acid in waters by the ignition of the water-residue with finely divided metallic copper.He also employed for the first time a process for the estimation of organic carbon which consisted in acidulating the water with sulphuric acid evaporating to dryness and then igniting the residue with cupric oxide as in an ultimate organic analysis. It is obvious however that this latter process could only make a distant approach to accuracy because the sul- phuric acid would not only expel volatile organic acids if pre-* Report to the President of the General Board of Health on the Metropolh Water Supply by Hofmann and Blyth. .t. Ann. Chemie u. Pharm cxxxii 215. VOL. XXI. G FRANKLAND AND ARMSTRONG sent but by liberating nitrous and nitric acids could scarcely fail to cause the oxidation and loss of other forms of organic matter.Nevertheless Welt z ien's memoir on the well waters of Carlsruhe is one of the most important contribntions ever made to this branch of analysis. In the following year W. A. Miller" gave an elaborate resume' of the processes which he considered to be most worthy of confidence together with some important modifications and valuable suggestions ; and in the same year Dr. Angus Smi t ht recommended and described certain modifications in the me of potassic permanganate in the examination of water for sanitary purposes. In the monthly examination of waters supplied to London one of us had frequently occasion to notice the serious imper- fections which still attached themselves to the best processes of water analysis hitherto employed for the determination of the organic matter and of the products of its decomposition.Since the autumn of 1866 we have been occupied with the critical examination of these processes and with attempts to place this branch of chemical analysis upon a more satisfactory basis. In laying the results of our inquiries before the Society it will be most convenient first to discuss the merits of the pro- cesses in general use up to the commencement of our investi- gation and then to describe the method of analysis which we now venture to recommend in dealing with the particular class of constituents above mentioned. The following is a list of the determinations which are usually made in the so-called partial analysis of potable waters and which have been submitted by us to examination :-1.Estimation of total solid constituents. 2. Estimation of organic and other volatile matter. 3. Determination of amount of oxygen required to oxidize the organic matter. 4. Estimation of nitrous and nitric acids. 5. Estimation of ammonia. We will examine these seriatim. I. Estimation of totul Solid Constituents.-This operation is usually performed by evaporating a given volume of the water Journ. Chem. SOC.,vol. xviii p 117. t Estimation of Organic Matter in Water by R. Angus Smith Ph.D. F.R.S. $c. 1865. ON THE ANALYSIS OF POTABLE WATERS. to apparent dryness with a known weight of sodic carbonate upon a steam bath ; and as the residue so obtained is generally employed for the determination of the organic and other vola- tile matters expelled on ignition it is dried at 120"-130° C.before being weighed. This process involves two errors in the first place the salts of ammonia are decomposed by the sodic carbonate the ammonic carbonate formed being expelled during evaporation; and secondly urea, if present in the water is slowly decomposed during evaporation with sodic carbonate ammonic carbonate being expelled. The loss of weight in the solid residue arising from the latter cause is Seen from the following determinations :-I. 1.505 grm. aodic carbonate and -038 grm. urea were dis-solved in 1litre of distilled water ; the soluti6n was evaporated on the steam-bath and the residue dried at 100°C.The latter weighed 1.526 grm. showing a loss of -017 grrn. or 44.7 per cent. of the urea employed. 11. *1grm. sodic carbonate and -05 grm. urea on being similarly treated lefi a residue which weighed *1205 grrn. showing a loss of -0295 grrn. or 59 per cent. of the urea present. It is difficult if not impossible entirely to avoid this loss but it is much lessened by omitting the sodic carbonate and drying the residue at 100' C. By this modification of the pro- cess the elements of water which would be expelled at 120"-130" C. are sometimes retained in the residue but as such water is in chemical combination it may be fairly aaid to belong to the solid constituents.It is only in the reddues of waters containing much calcic and magnesic sulphates and chlorides that the weight of the elements of water so retained is con-siderable. Its small amount in the case of Thames water is seen from the following determinations made upon three different samples :-Solid residue Solid residue dried at dried at 100" 12O0-l3O0 C. 100,000parts of Thames water gave 27.02 pts. 26.54 100,000 26.70 , 26.20 9 ? 100,000 26-10 , 26-02 9 79 IT. Estimation of Organic and other Volatile Matter.-Thie deter-mination is effected by the gentle ignition in contact with air (32 80 FRANKLAND AND AR3fSTRONG of the solid residue obtained by evaporation and subsequent drying at 120°-130° C. as above described After being allowed to cool the residue is repeatedly treated with a saturated solu- tion of carbonic anhydride in water until on being again dried at 120'-130° it ceases to gain weight.The loss in weight experienced by a water residue on incineration although now no longer regarded as an exact ponderal expression of the organic matter present in the original water is yet considered by many chemists to afford at least an approximate indication of the amount of organic impurity. How far any reliance can be placed upon it in this respect may be judged of by the consideration of the following sources of error which we have found in the examination of this process. We have already shown that urea when present in water contaminated with sewage is partially dissipated during evaporation with sodic carbonate ; thus a portion of the organic impurity of a water may be lost before the determination of the '' organic and other volatile matter " is made but this error is still further increased in the case of urea and probably also in that of other nitro- genous organic compounds by the impossibility of expelling the whole of the organic matter on ignition as is seen from the following experiments :-I.A water-residue consisting of 0997 grm. sodic carbonate and 00205 grm. urea lost on ignition 0003grm. 11. A water-residue consisting of 1.542 grm. sodic ca,rbonate and -039grm. urea lost on ignition *011grm. 111. A water-residue consisting of 1.505 grm. sodic carbonate and 0038 grm. urea lost on ignition ,016 grm.The results of these experiments may be thus expressed :-I. 11. 111. Percentage of organic matter expelled on ignition 14.6 28.2 42.1 39 99 99 left in residue 85.4 71.8 57.9 It is probable that the organic matter left in the residue is in the form of sodic cyanate or cyanurate. Another source of error which we have repeatedly encountered in the use of this process consists iii a continued increase of weight by successive treatments of the incinerated residue with an aqueous solution of carbonic anhydride until in some ca8es ON THE ANALYSIS OF POTABLE WATERS. the weight of the incinerated residue even exceeds that which was observed before incineration. This remarkable pheno-menon does not arise from any solid residue left by the solution of carbonic anhydride itself because this source of fallacy was carefully eliminated and moreover it occurs only with par- ticular samples of water but with these samples it is always observed when the determination is repeated.It occurs also in an equal degree when a solittion of ammonic carbonate is substituted for one of carbonic anhydride. It is difficult to account for this increase in weight but when it occurs the determination of the loss by ignition becomes impossible because the analyst does not know when to discontinue the treatment of the residue with solution' of carbonic anhydride. These facts show how difficult it is to interpret the meaning of the loss on ignition experienced by a water-residue; it may arise entirely from organic matter or it may be exclusively due to the dissipation of mineral ingredients.On the one hand there may be much more organic matter in a water than is represented by the total loss on ignition indeed; we have not unfiequently observed in the analysis of waters greatly con- taminated with sewage that the loss on ignition has actually been considerably less than the weight of the carbon alone contained in the organic matters. On the other hand this determination may indicate the presence of a considerable amount of organic matter in a water which is wholly free from it. All that can be inferred fi-om the loss on ignition is that when it is large the water is probably contaminated with animal or vegetable organic matter or has been previously in contact with decaying animal matters.III. Determination of amount of Oxygen necessary to Oxidise t7~e Organic Matter.-The uncertainty which surrounds the deter- mination of the organic and other volat,ile matter by the igni- tion of the dried water-reaidue has led to attempts being made to estimate indirectly by means of potassic permanganate the amount of organic matter present in the water before evapora- tion. Potassic permanganate when dissolved in water readily yields oxygen to many substances capable of combining with this element; thus if it be added to water acidulated with sulphuric acid and containing oxalic acid in solution the latter is completely and rapidly converted into carbonic anhydride and water at the expense of oxygen derived from the permsn- FRANKLAND AND ARMSTRONG ganate; and it is found that eight parts by weight of oxalic acid in being thus oxidised abstract almost exactly one part by weight of oxygen from the permanganate the latter being converted into manganic sulphate.In undergoing this chemical change the rich violet colour of the solution of potassic perman- ganate vanishes and it is thus easy to ascertain by the non- disappearance of the characteristic tint of the permanganate when the oxidation of the oxalic acid is complete. Now a similar disappearance of colour occurs when the solution of potassic permanganate is added to an acidulated sample of potable water containing organic matter and it has been assumed that as in the case of the oxalic acid the organic matter contained in the water is completely oxidised by the permanganate which is thus thought to indicate the amount of oxygen required for this purpose.Dr. Letheby has even employed this reaction for the estimation of the actual weight of organic matter contained in a known volume of water on the assumption that every eight grains of organic matter contained in a sample of water rob the permanganate solution of one grain of oxygen. Such a method of ascertaining the actual amount of organic matter in a water or even the amount of oxygen required to convert this organic matter into its final products of oxidation would be invaluable on account of the extreme facility with which it can be applied; but unfortunately the hrther study of this process reveals its utter untrust-worthiness.By the addition of known weights of different organic sub-stances to equal volumes of pure distilled water the latter was artificially contaminated with a known proportion of each kind of organic matter. Every sample of water so contaminated was made to contain three parts of organic matter in 100,000. The amount of oxygen which this organic matter abstracted &om the potassic germanganate was first carefully ascertained and then the actual amount of organic matter present in the water was calculated on the assumption that eight parts by weight of organic matter consumed one part by weight of oxygen from the permanganate. The Same test was also applied to another sample of distilled water from which all organic matter was carefully excluded but to each 100,000 parts of which three parts of sodic nitrite were added.The importance of the last experiment will be evident when it is ON TElE ANALYSIS OF POTABLE WATERS. remembered how frequently nitrites are present in potable waters. The amount of oxygen consumed was determined for two different periods of time viz. :-First for a period at the end of which the acidulated and Contaminated water remained tinted with permanganate for ten minutes after the addition of the latter ; and secondly for a period of six hourB during the whole of which time the permangafiate was present in excess. The results are contained in the following table where they are compared with the known amount of organic matter pre- sent and the known amount of oxygen which that organic matter would require for its complete oxidation :-I Name of Substance 3 parts of which were contained in 100,000 F && parts of water.d2 aa Gum arabic.. .......... 00102 -0350 3 *55 *083 *280 3.0 Cane sugar.. .......... *0064 ,0152 3 -37 -051 *Ill 3.0 Starch ................ -0143 *0302 3.55 '114 *241 3-0 Gelatin .............. *0792 -1836 6 *'76 -634 1.469 3-0 Crestin .............. -0080 -0172 6 -59 -064 -138 3.0 Alcohol .............. -0093 -0164 6 .26 -074 -131 3.0 Urea. ................. -0092 -0119 6 .40 -074 ,095 3.0 Hippuric acid.. ........ -0328 *0600 5 -90 -262 -480 3.0 Oxalic acid (crystallised) *3747 *3750 *38 2 -998 3.000 3.0 Sodic nitrite ..........-6910 -6913 0 so0 5,521 5.530 0.0 From this table it is seen that of the nine kinds of organic matter operated upon only one was completely oxidized by potassic permanganate even after the lapse of six hours; whilst it will be remarked that urea hippuric acid and crea-tin-three organic substances likely to be present in water recently contaminated with sewage-suffer an oxidation which even in the most favourable case only reaches &th of com- plete oxidation; whilst if the attempt be made to calculate the amount of these substances present in the water from the quantity of oxygen so absorbed instead of finding three parts of each in 100,000of water we obtain only-138 part of creatin FRANKLAND AND ARMSTRONG -095 part of urea and -480 part of hippuric acid.011 the other hand the mineral salt sodic nitrite weight for weight sur-passes every form of organic matter experimented upon in the avidity with which it absorbs oxygen; and three parts of this inorganic substance in 100,000 of water would actually by the mode of calculation above described indicate no less than 5+ parts of organic matter. Thus it is evident that for the estimation of the amount of organic matter in water or the quantity of oxygen necessary to oxidise that organic matter permanganate of potash is utterly untrustworthy. The fallacy of the permangarlate test has often been suspected but it was imagined that although not to be relied on for quantitative esti- mations yet its rapid decoloration afforded positive evidence of the presence of organic matter in actual putrescence and con- sequently in its most dangerous condition.We fear however that even for this subsidiary purpose the permanganate is not sufficiently trustworthy. Fresh urine contains no organic matter in a state of putrescence but even when largely diluted it decolorises the permanganate with almost the same rapidity as potassic nitrite. With all these defects however this re-agent may still be used in certain cases as a qualitative test where there is 110 opportunity for accurate analytical examina- tion. Thus if a clear and colourless water decolorises much of the permanganate solution the water ought to be rejected for domestic use as being of doubtful quality; for although such a water may be absolutely free from nitrogenous organic impurity yet its decolorising action upon the permanganate would indicate with considerable certainty that it had been in contact with decaying animal mat'ters.Should the water how- ever instead of being colourless be tinged of a yellow or brownish yellow colour when viewed through a considerable stratum as in a quart decanter for instance its capability of decolorising a considerable amount of permanganate solution ought not to be regarded with the same suspicion as a similar reaction with a colourless water because the yellow tint of auclz waters is generally owing to the.presence of peaty matter which though innocuous has the power of decolorising potassic permanganat e.The depth of colour which a sample of water exhibits when viewed through a stratum some two feet in thickness,-has also been regarded aB an indication of the amount of organic matter 85 ON THE ANALYSIS OF POTABLE WATERS. contained therein. It appears to be so regarded for instance by DF. Letheby who thus speaks of the use of this test :-“The oxidizable organic matter is determined by a. standard solu- tion of permanganate of potash-the available oxygen of whicl is to the organic matter as 1is to 8; and the results are con- trolled by the examination of the colour of the water when seen through a glass tube two feet in length and two inches in diameter.” * The tinctorial power of many colouring matterg is so great as to render them distinctly a.ppreciable to tbe eye when their amount is far too minute to be detected gravimetrically ; thus a litre of water distinctly tinted with ink or magenta contains an amount of either of these colouriiig matters too small to be appreciated by the most delicate balance.The yellowish or brownish colour of water appears also to be of this character for it may be removed completely by agitation with aluiiiinic hydrate and yet a considerable amount of organic matter is still left in the water. Thus a sample of water which had been 80 treated and which exhibited the blue-green tint of distilled water when viewed through a thick stratum still coiitained 0160part of organic carbon in 100,000 parts.It is therefore evident that no reliance can be placed upon colour as an indi- cation of the amount of organic matter in waters for although a dark-tinted water probably contains a considerable amount of organic matter it by no means follows that a colourless water may not contain even a larger proportion. Thus a sample of peaty water possessing a decided brown tinge contained but -256 part of organic carbon in 100,000 parts whilst a sample of water from North Wales which had been in contact with the fine mud of the stamping-engines of mines was perfectly colourless although it contained no less than -544 part of organic carbon in 100,000 parts. IV. Estimation of Nitrous and Nitric Acids.-The best process hitherto employed for this estimation is that proposed by the late Dr.Pugh. It depends upon the conversion of stannous chloride into stannic chloride in the presence of fiee nitric acid whilst the latter is transformed into ammonia. The application of this process to water analysis has been described by Miller.? * Dr. Letheby’s reports on the Metropolitan Waters to the Association of Medical Officers of Health. t Journ. Chem. SOC.,vol. xviii p. 117. 86 FRANKLAND AND ARMSTRONG If nitrites be also present they may be converted into nitrates by the cautious addition of potassic permanganate to the water previously slightly acidified. The process is easy of execution and extremely delicate ;but unfortunately a@ Messrs. Chapman and Schenk have shown stannous chloride is converted into stannic chloride by many organic substances containing oxygen such a8 starch sugar &c.The following experiments prove that this effect of starch and sugar takes place to such an extent as to render the process entirely untrustworthy. I. *lo0grm. starch was digested for 20 ininutes in a sealed tube with 3 C.C. of stannous chloride solution at a temperature of 17OOC. Before digestion 3 C.C. of the same solution of stan- nous chloride required 17-85 C.C. of a standard solution of dipotassic dichromate to oxidize it. After digestion it required only 14.1 C.C. The oxidizing action of the starch was equiva- lent to that of *00375 grm. of N20,. 11. -100 grm. sugar similarly treated at 15O0C. oxidized an amount of stannous chloride equivalent to 7-0C.C.of the standard solution of dipotassic dichromate corresponding to -007grm. of w5. 111. a100 grm. starch similarly treated at 150°C. oxidized an amount of stannous chloride equivalent to 4.2 C.C. of the standard dipotassic dichromate solution corresponding to *0042grm. of N2O,* IV. ~100grrn. atarch digested at 12OoC. oxidized an amount of stannous chloride equivalent to 2.5 C.C. of the standard di- potassic dichromate solution corresponding to -0025 grm. of w5= V. The last experiment repeated with *I grm. of sugar instead of starch at 120°C. gave results corresponding to *0058grm. of N,05. Experiments 11,111,IV and V were made in order to ascer-tain whether the oxidizing action of sugar and starch could not be prevented by operating at lower temperatures but they show that these substances still oxidize very powerfully even at a temperature below the minimum required according to Pugh for the performance of his reaction.V. Estimation of Armmonia.-The determination of ammonia in potable waters is usually made by rendering the water alkaline either by baric hydrate or sodic carbonate and then distilling off about one-fourth of its volume. In the diatillate ON TWE ANALYSIS OF POTABLE WATERS. the ammonia is then estimated either by neutralisation with a standard solution of dilute acid or by Hadow’s modification of Nes sler’s reaction. In its application to waters recently con- taminated with sewage this process is liable to considerable inaccuracy owing to the gradual production of ammonia when an alkaline solution of urea is boiled.Thus the ammonia found exceeds that originally contained in the water. This error haa already been pointed out by Mr. Chapman who recomniends that the ammonia determination should be made by the appli- cation of Nessler’s solution directly to the water. We find however that the yellowish colour of many potable waters presents a formidable obstacle to success unless the wakr be first decolorised as we recommend below ; besides waters con- taining chalk in solution become turbid on the addition of the Nessler test and any turbidity is utterly fatal to accuracy in this determination. Having thus pointed out the inaccuracies which attach themselves to the usual determinations in a water analysirj we will now describe the processes which we propose as substi- tutes for or modifications of those which have been hitherto employed.They may be thus enumerated :-1. Estimation of total solid constituents. 2. Estimation of the carbon and nitrogen contained in the organic portion of the solid constituents (organic carbon and nitrogen). 3. Estimation of nitrogen in the form of nitrates and nitrites. 4. Estimation of ammonia. 1. Estimation of total Solid Constituents.-Half a litre of water is evaporated to dryness as rapidly a8 possible in a weighed platinum capsule on a steam- or water-bath; after drying the residue at 100’ C. the capsule is again weighed.We have already given our reasons for the non-addition of sodic carbonate to the water before evaporation and also for drying the residue at 100’ instead of 120’-130’ C. As we propose to abolish al- together the fallacious estimation of ‘‘ loss on ignition,” the retention of the elements of water in this residue is of no moment; they always exist there in the solid condition and are hence quite legitimately included amongst the solid consti- tuents. 2. Estimation of Organic Carbon and Nitrogen.-No process has yet been devised by which the amount of organic matter in FRANKLAh'D AND ARMSTXONQ water can be even approximately estimated but we have now to describe a method by which the two most important elements -carbon and nitrogen-can be determined with considerable accuracy.The estimation of the organic carbon in a water containing both carbonates and carbonic anhydride in solution is as might be anticipated an operation of more than ordinary difficulty. It is obviously necessary in the first place to expel both com- bined and dissolved carbonic anhydride and this must be done in such a manner as to prevent the organic matter from being subject to the oxidising action which would necessarily result from the liberation of nitric and nitrous acids which are pro- bably never entirely absent from potable waters. We en- deavoured to effect this by the addition of boric acid to the water during evaporation. Bloxam has shown" that for the expulsion of one molecule of carbonic anhydride from alkaline carbonates six molecules of boric acid are necessary.Approxi-mately this appears to be true also of the carbonates of the alkaline earths ;nevertheless after the addition of six molecules of boric acid to each molecule of combined carbonic anhydride we still found an amount of carbonate in the residue which though small was sufficient seriously to vitiate the result of the subsequent determination of organic carbon. After many abortive attempts to overcome this difficulty we found in sul-phurous acid a re-agent which not only completely expels carbonic anhydride from the water but also permits of the simultaneous determination of organic nitrogen with great accuracy by completely removing during evaporation every trace of the nitrogen existing in the form of nitrates and nitrites and thus leaving in the dry residue the organic nitrogen associated only with one remaining nitrogenous body viz.ammonia. For the successful application of sulphurous acid to this pmpose it is not sufficient to add an excess of this acid to the potable water and then evaporate to dryness since under these circirmstances traces of carbonates are always found in the residue. It is in fact necessary after the addition of excess of sulphurous acid to boil the water for two minutes in order to insure the complete expulsion of carbonic anhydride before evaporation on the steam-bath begins. If this precaution be * Jwm. Chem. SOC.,vol. xii p. 177 ON THE ANALYSIS OF POTABLE WATERS. observed numerous experiments have shown that no trace of carbonic anhydride is evolved on adding hydrochloric acid to the dry residue.By availing ourselves of that admirable instru- ment the Sprengel-pump we are able to combine in one opera- tion the determination of carbon and nitrogen ill a water-residue by an analytical process of such simplicity and extreme delicacy that we believe it will be found generally useful in the analysis of all organic compounds containing nitrogen which are not volatile at ordinary temperahres. By this process *000001grm. of nitrogen and *0000005grm. of carbon are distinctly measur- able quantities. The following ia the mode of conducting this operation :-As soon as possible after the collection of the sample of water 2 litres are poured into a convenient stoppered bottle and 60 C.C.of a recently prepared saturated solution of sulphurous acid are added. Should the water contain oxidizable or putrescible organic matter this addition of sulphurous acid promptly arrests any further change and the remaining operations may now be conducted at leisure. One-half of this sulphurized water is now boiled for two or three minutes," and unless it contained a con-siderable amount of carbonates -2 grm. of sodic sulphite is to be added during the boiling so as to secure the saturation of the sulphuric acid formed during the subsequent evaporation. To secure the expulsion of the nitrogen existing as nitrates it is also desirable to add a couple of drops of solution of ferrous or ferric chloride.The boiled water is then evaporated to dryness in a hemispherical glass capsule of about 100C.C. capacity upon a steam-or water-bath care being taken to keep the capsule well covered with a disc of filter paper stretched over a light cane hoop and also to preserve the atmosphere of the room in which the operation is performed as free from ammonia as pos-sible. If the first-mentioned precaution be neglected the ac- cess of floating particles of dust during the evaporation will introduce a considerable error into both carbon and nitrogen determinations. At one period of our investigations we feared that it would be necessary to perform the evaporation in vacuo but we prove below that this would be an unnecessary precau- tion for if the evaporation be conducted under paper the * In operating with waters strongly contaminated with sewage it is desirabIe that the flask in which this operation is performed should be furnished with an in-verted Liebig's condenser in order to prevent the Ioas of volatile organic con.stituenls and ammonia. With all ordinary waters this precaution is unnecessary FRANKLAND AND ARMSTRONG amount of nitrogen introduced by atmospheiic dust and am-monia (and this in the worst of all possible localities the middle of London) only amounts to a maximum of -00002 grm. per litre of water evaporated consequently when this is deducted &om the amount of nitrogen actually obtained by combustion the residual ei-ror is almost a vanishing quantity.The process of evaporation under a paper cover fitting tightly upon the edge of a glass dish without a Zip,is in fact one of difision in which the atmospheric air constantly being exchanged for aqueous vapour is filtered through a porous diaphragm. The evaporation being completed and the glass capsule placed upon a sheet of glazed paper a few grammes of powdered plumbic chromate are to be introduced and gently triturated with the dry residue by means of an agate or glass pestle. When the mixture * has been made as perfect as possible the contents of the capsule are to be transferred to a combustion tube about 16 inches long and sealed at one end the capsule is then ringed two or three times with fresh quantities of chromate which are also transferred to the cornbustion tube.The latter is then charged in the usual manner with granu- lated cupiic oxide and about 3 inches of bright copper turn- ings. The open end must now be drawn out before the blow- pipe as shown in figure I and the tube being laid in a FIG. 1 combustion-furnace the drawn out extremity is to be connected with a Sprengel-pump by means of a piece of india-rubber ON THE ANALYSIS OF POTABLE WATERS. tubing care being taken that the extremities of the two glass tubes touch each other or nearly so within the caoutchouc connector. The latter being then plunged beneath water in the vessel A and the hrnace around the froni part of the com- bustion-tube lighted the pump is to be worked until the tube is exhausted as completely as possible an operation which requires from five to ten minutes.* The flow of mercury is then stopped.The recurved delivery end of the pump b dips into a mer- curial trough C andan inverted tube d filled with mercury is placed over it in a convenient position for receiving the gaseous products of the ignition. The combustion must now be con-ducted in the usual manner care being taken that when the organic matter begins to burn the operation proceeds very slowly until the vacuum becomes considerably impaired ; otber-wise traces of carbonic oxide may be produced. A combustion usually lasts from three-quarters of an hour to an hour ; at its conclusion unless the water-residue contained much organic matter no gas will have passed into the inverted tube.The pump is again set to work and in from five to ten minutes the whole of the gases will be transferred into the vessel placed for t,heir reception. Unless the heat of the furnace be excesaive the combustion-tube will rarely collapse; but if it should do so the metallic copper and granulated cupric oxide support the glass and prevent any obstruction to the passage of the gases. In many scores of combustions made by this process no single instance of vitiated result has occurred from this cause. The gases collected consist of carbonic anhydride nitric oxide and nitrogen. The separation and determination of these by well-known methods is exceedingly simple and in a manometric gas apparatus such as that described by one of US,^ is the work of a few minutes only.* As it is obviously necessary that the leakage of atmospheric air into the pump should be rendered impossible the caoutchouc pinch-cock at B should be enclosed in a wide piece of vulcanized tube the annular space between the two tubes being filled with glycerine. The clamp is placed outside both tubes. The wide piece of tube is fastened upon the glass tube below the pinch-cock by the aid of an india- rubber cork whilst it terminates Considerably above the joint ;the interior caout- chouc-joint is therefore entirely immersed in glycerin and all possibility of leakage of air entirely prevented. We find in fact that by this arrangement the vacuum is still perfect after the lapse of several days. The calibre of the pump-tube which we prefer is one millimetre and it is adrisable to allow the mercury to flow very slowly until the exhaustion is nearly complete when a rapid stream is necessary to expel the remaining traces of air or gas.t Journ. Chem. Soc. vol. vi p. 197. FRANKLAND AND ARMSTRONQ A simplified form of this apparatus designed especially for the examination of all gaseous mixtures incident to water analysis is described in the memoir immediately following the present where the method of analysing these mixtures is also given. The weights of carbon and nitrogen contained in the carbonic anhydride nitric oxide and nitrogen gases having been deduced fiom the respective volumes of these gases the numbers so obtained are expressed in parts of these elements con- tained in 100,000 parts of' the water.The nitrogen thus found may have been present in the water first as a constituent of organic matter (organic nitrogen) and secondly as a constituent of ammonia. The latter if present is determined in the original water by N ess1 e r ' s test as described below and the nitrogen existing in this form being deducted from that obtained on combustion gives the amount if any of organic nitrogen present. It is obvious that the accuracy of this method of combustion will depend in a great measure upon the perfection of the vacuum obtained by the Sprengel pump. In order to ascertain the error due to this cause the following experiments were made :-I. -01 grm. sugar was burnt in the Bame way as a water residue.After absorption of carbonic anhydride there remained 0019 C.C. of nitrogen at 0' C. and 760 mm. pressure. 11. 001grm. sugar similarly treated gave *013C.C. of nitrogen at Oo C. and 760 mm. pressure. If these numbers be referred- to the residue of one litre of water (the quantity usually operated upon) the excess of nitrogen due to the imperfection of the Sprengel vacuum would be- . I. ,0024 part of nitrogen in 100,000parts of water. 11. -0016 part of nitrogen in 100,000 parts of water. It will be seen that this error which includes also any nitrogen retained or occluded in the cupric oxide &c. is very insignific'ant; nevertheless it would be necessary to allow for it if it were not included in another correction which consists in evaporating a lit're of distilled water,* acidified as usual This distilled water should be previously purified by toiling for 24 hours with alkaline potassic permanganate.It should then be distilled the first portions of the distillate being rejected so long as t.hey show any reaction with Xessler's test. Finally this distillate should be slightly acidified with sulphuric acid and rectified. ON THE ANALYSIS OF POTABLE WATERS. with 15 C.C. of sulphurous acid and containing about 01pa of recently ignited sodic chloride. The residue fiom this water must now be burnt in vacuo in the usual manner and the carbon and nitrogen deducted fi-om the amount of these elements obtained fiom the residues of other watera submitted to analysis.It is advisable that each analyst should perform several blank operations of this kind so as to be able accurately to correct for the combined errors of his own manipulation and apparatus. In our own case we find these errors on the average of four blank analyses to amount to-Carbon *00032grm. in 1litre of water. Nitrogen *000045grm. in 1litre of water. It is scarcely necessary to add that to insure a minimum in these errors it is of the utmost importance carefully to guard against every access of organic matter and especially of nitro-genous compounds to the water and the substances used in the analysis. Cupric oxide prepared from the nitrate should on no account be used since even after being actually fused it evolves considerable quantities both of carbonic anhydride and nitrogen when ignited in vacuo.The oxide must be made by igniting sheet copper in a current of air in a muffle or other convenient apparatus. This oxide in a coarsely granular or scaly condition should be at once transferred to a stoppered bottle over the neck of which a small beaker is inverted to protect it from dust. The fiised plumbic chromate should be heated to redness with frequent stirring for a couple of hours and then carefully transferred to another bottle similarly pro- tected. As these substances do not require to be either dried or ignited again before uae they should be transferred as required from their respective bottles direct to the capsules or com-bustion tubes and any portion of them once removed from the bottle should on no account be returned there without being first ignited for two hours.The extent to which this method can be depended upon for the determination of the minute amounts of carbon and nitrogen contained in a water residue was tested by the following experi- ments :-I. *0352 grm. sugar was dissolved in one litre of distilled water together with about 05 grm. of sodic carbonate previousIy VOL. XXI. H FRANKLAND AND ARBISTRONGI. converted into sulphite 15 C.C. of a sa.turated solution of sub phurous acid were then added and the liquid boiled for three minutes. The residue left after evaporation to dqpess on the steam-bath gave on combustion an amount of carbonic anhy-dride coi-responding to 001463 grm.carbon. 11. 00347 grm. sugar similarly treated gave *01386 grm. carbon. 111. 00114 grm. sugar similarly treated gave -00440 grm. carbon. IV. 00122 grm. sugar similarly treated gave -00530 grm. carbon. V. *0115grm. sugar 00094grin. ammonic chloride and -8 grm. sodic carbonate (previously converted into sulphite) treated in like rrianner gave *004344 grm. carbon and -0025415 grm. nitrogen. VI. *010 grm. urea and *8 gm. sodic carbonate similarly treated gave -0017704grm. carbon and a00463 grm. nitrogen. VII. *01025grm. urea and 43 grm. sodic carbonate treated as before gave a00211 grm. carbon and -00357 grm. nitrogen. VIII. 00104grm. urea and -8 grm. sodic carbonate similarly treated gave -0033865 grm. carbon and *004675grm.nitrogen. IX.*02O2 grm. urea and one litre of solution of dihydric calcic dicarbonate boiled with 15 C.C. of sulphurous acid solu- tion and evaporated gave -00452 grm. carbon and -00887grm. nitrogen. X. -025 grm. hippuric acid and a5 grm. sodic carbonate (con- verted into sulphite) dissolved in one litre of water boiled with 10 C.C.of sulphurous acid solution and evaporated to dryness gave -01386grm. carbon and -00203 grm. nitrogen. Expressed in parts per 100,000 of water evaporated the following are the results of these experiments :-Calculated. Found. No. 11 Organic carbon .. .. 1-482 1-463 1, No. 11. ........ 1.460 1.386 No. 111. 9 ........ 0480 0440 79 No. IV. ........ *514 -530 ........ -484 -434 ” No. V. {Nitrogen ............ -246 -254 No.vI. {Organic carbon ...... 0200 -17’7 .. nitrogen .... *466 -463 OW THE ANALYSIS OF POTABLE WATERS. 95 Calculated. Found. No. {Organic carbon ...... -205 -211 , nitrogen .... -478 *357 , carbon ...... 908 -239 , nitrogen .... 0484 -468 , carbon ...... ,404 0452 No. IX. { , nitrogen .... *942 -887 No. x. { , carbon ...... 1.508 1.386 , nitrogen .... *195 -203 When it is considercd that these results were obtained from very minute amounts of the respective organic matters which were first dissolved in a large quantity of water and then re-covered by evaporation and further that some of the organic substances experimented upon are exceedingly prone to change the correspondence of the experimental with the calculated numbers is as close as could be anticipated.The following results obta,ined with actual waters also tend to inspire confi- dence in this method of analysis :-XI. Two litres of the same sample of water were succes- sively analysed five days apart. They gave the following amounts of organic carbon and nitrogen in 100,000parts :-No. I. No. 11. Organic carbon ................ 1.930 1.010 Organic nitrogen .............. *I98 *207 XII. Three mixtures of sewage and distilled water were made in such proportions that 1 litre contained respectively 100 c.c. 10 c.c. and 1C.C. of sewage. Some solution of dihydric calcic dicarbonate was added to the second and third to form a tangible residue. They were then treated with sulphurous acid in the manner above described and evaporated to dryness.Their residues gave on combustion the following results per 100,000parts of water evaporated :-100 C.C. sewage. 10 C.C. sewage. 1 C.C. sewage. 900 C.C. water. 990 C.C. water. 999 C.C. water. Organic carbon in of1 100,000 part8 -302 *033 *005 the mixture .... Organic nitrogen and nitrogen of} *330 -033 -004 ammonia. ....... H2 E'RANKLAND AND ARMSTRONG It has been already stated that the nitrogen obtained on the combustion of a water-residue is made up of the organic nitrogen plim the nitrogen of any ammonia that may have been contained in the water but that it includes no trace of the nitrogen which may have been present in the form of nitrates and nitrites the latter having been completely destroyed during the evaporation with excess of sulphurous acid.Such an ex- pulsion of the nitrogen of nitrates and nitrites is a remarkable reaction and could scarcely have been predicted; indeed it takes place to a very partial extent only when a nitrate is dis- solved in water and evaporated with excess of sulphurous acid in imitation of a natural water ; neither is the result very dif- ferent when sodic chloride or calcic or magnesic carbonate is added. Thus the residue from half a litre of distilled water to which had been added *05grm. potassic nitrate (= 0007 pm. nitrogen) *OOO1 grrn. ammonia -1 grm. sodic chloride and 15 C.C. of a saturated solution of sulphurous acid yielded -00161 grm. nitrogen. One litre of distilled water containing *Igrm.sodic chloride 01grm. potassic nitrate (= -014 grm. nitrogen) one drop of a strong solution of soluble glass and 15 C.C. of a saturated solu-tion of sulphurous acid treated like a natural water yielded on combustion -00222 grm. nitrogen. One litre of distilled water containing -1grm. sodic chloride -1grm. potassic nitrate and 15 C.C. sulphurous acid solution similarly treated gave *00259 grrn. nitrogen. The presence of a minute amount of iron or of a phoBphate reduces to zero the amount of nitrogen retained from nitrates. Thus 1 litre of distilled water -1grm. sodic chloride *Ignn. potassic nitrate (= -014 grm. N.) 2 drops of a moderately con- centrated solution of hydric sodic phosphate and 15 C.C. of sulphurous acid solution gave no nitrogen on combustion of the solid residue.Half a litre of distilled water containing 01grm. potassic nitrate and 2 drops of a solatim of ferric chloride evaporated with 10 C.C. of sodic sulphite solution and 15 C.C. of a saturated solution of sulphurous acid gave no nitrogen on combustion. One litre of distilled water containing -1grm. sodic chloride -1grm. potassic nitrate 1 drop of solution of ferric chloride and 15 C.C. of sulphurous acid sohition gave no trace of nitrogen on combustion and the same result was obtained in a duplicate ON THE ANALYSIS OF POTABLE WATERS. experiment. Three drops of a solution of ferric chloride also removed all traces of nitrates from half a litre of a natural water when evaporated in vacuo although the water coiitained no less than 2-466 parts of nitrogen as nitrates and nitrites in 100,000parts.The nitrogen was also completely expelled during the evapo- ration of an artificial water to which the following ingredients were added :-*Ol grm. magnesia -1 grm. calcic carbonate 01grm. sodic chloride *O1 grm. potassic chloride 1 drop of solution of soluble glass 1drop of solution of ferric chloride 2 drops of solution of hydric sodic phosphate *Igrm. potassic nitrate and 15 C.C. of sulphurous acid solution. There is probably no natural water containing an appreciable quantity of nitrates or nitrites which does not also contain either iron or phosphoric acid; nevertheless it is advisable to add one or two drops of ferrous or ferric chloride to the portion of water which is evaporated for combustion in order to place beyond the possibility of doubt the complete expulsion of the nitrogen of nitrates and nitrites.Since we began to use this process for the estimation of organic carbon and nitrogen in waters Messrs. Wan klyn Chapman and Smith have proposedanew method for the determination of the latter element in potable waters. Their process is founded upon a highly remarkable change which albumin and some other organic substances undergo during prolonged ebullition with an alkaline solution of potassic per- rnanganate by which their nitrogen is converted into ammonia. Unfortunately however this conversion is never complete ; neither is there any guarantee that all the different forms of nitrogenous organic substances in water will thus yield up their nitrogen in the form of ammonia.That some such substances do not thus evolve their nitrogen when submitted to this pro- cem is evident from the following results obtained with three bodies taken at random from a collection of chemicals :-I. -01 grm. strychnine dissolved in 1 litre of distilled water (not previously purified) and distilled nearly to dryness with caustic potash and potassic permanganate gave *00032 grm. ammonia. 11. 002 grm. narcotine similarly treated gave *000312 grm. ammonia. 111 -02grm. quinine sulphate gave *000728 grm. ammonia FRANKLAND AND ARMSTRONG The following comparison of the amounts of ammonia actually obtained with those which ought to be yielded by the weightg of the respective substances operated upon shows that in each case a large proportion of nitrogen was not evolved as am- monia :-Ammonia evolved.Calculated. Found. Strychnine .............. *031Ol grm. *00032 grm. Narcotine. ............... *00068 , *000312 9y Quinine sulphate ........ *00128 , *000728 9y We have also tested the permanganate process by applying it to a form of nitrogenous organic matter which is very fie- quently met with in natural waters viz. peaty matter. Some peat collect'ed by one of us -from Leyland moss at a depth of three feet below the surface and placed immediately in a well-corked glass vessel was digested at 100" C. for a couple of hours in distilled water rendered slightly alkaline by caustic soda.100 C.C. of the dark-coloured liquid so obtained was made up to one litre with distilled water and after the determination of ammonia by ebullition with sodic carbonate was submitted to the permanganate process so long arJ am-monia was evolved. Another 100 C.C. of the same liquid was acidified with sulphurous acid boiled for two minutes then evaporated to dryness in vacuo and the dry residue submitted to combustion in vacuo. The following amounts of organic nitrogen per 100,000 parts of liquid were obtained :-Permanganate Process. Combustion Process. 0308 1-015 Another portion of the same liquid was acidified with dilute sulphuric acid ; the copious brown precipitate which separated was collected on a filter and after being dried at 100"C.was reduced to fine powder. Two separate ceutigrms. of this pre- cipitate were respectively submitted to the permanganate and combustion processes. Two equal volumes (100 c.c.) of the filtered liquid were also reHpectively treated by the two pro- cesses the portion used for combustion being evaporated under paper upon a steam-bath. The ammonia was determined in this liquid a8 usual. The following are the amounts of organe'c. nitrogen obtained :- ON TKE ANALYSIS OF POTABLE WATERS. Permanganate procem. Combustion proceaa 001grm. of peat precipi- tate yielded of organic *000052grm. *0001138 grm. N. nitrogen . . . . . . . . . . 100,000 parts of filtrate born peat precipitate i -108 gave of organic nitro- -291 gen .. . . . . . . . . . . ,. 200 C.C. of another sample of peat solution treated by the two processes yielded the following amounts of organic nitrogen per 100,000 parts of liquid :-Permanganate process. Combustion process. -422 1.175 Two separate litres of an artificial water made by diflksing some peat in distilled .water for several days (without the addi- tion of alkali) and then filtering were treated by the two pro- cesses and yielded the following amounts of organic nitrogen per 100,000 parts of water :-Permanganate process. Combustion process. *022 *076 The extension of this comparison of the two processes to natural waters confirms in a large majority of cases the con-clusion which is forced upon ug by the above experiments viz.that nitrogenous organic substances do not uniformly yield up the whole or nearly the whole of their nitrogen in the form of ammonia when boiled with alkaline potassic permanganate ; indeed W anklyn has recently discovered that even in regard to albumen itself his first statement in reference to this point requires modification and he now states” 66 The albuminoid ammonia’ is not the total amouiit of amnionia which the albumen is capable of giving but appears to be two-thirds of the total quantity being at any rate a constant firtction of the total quantity.” Neither the above nor the following results show either that two-thirds of the total nitrogen is evolved in the shape of ammonia or that the fiaction of the total nitrogen evolved in the permanganate process is a con-stant one.We have tested the two processes side by side upon more than 100 different samples of natural waters and we * Joarn. Chcm. SOC. vol. xx p. 593. FRANKLAND AND ARMSTRONG find that as a rule to which however there are Rome excep-tions the permanganate process gives results considerably below those obtained by combustion as in the following cases :-Organic nitrogen in 100,000 parts Organic nitrogen in 100,060 parts of water. of water. By permanganate By combustion. By permanganate By combustion. procees. process. -006 .... -010 -002 .... 410 0006 .... -011 -002 .... *008 *006 .... -010 -003 .... 0008 *002 .... -011 *016 .... 0068 0016 .... -042 *003 .... 0006 *002 .. .. -009 *001 .... -012 0006 .... .022 *002 .... *011 -000 .... *007 *ooo ,... -007 -013 .... 0043 *011 .... *058 *012 .... -027 *024 .... -061 -006 .... -031 *030 .... 0062 In Borne czmes where as a rule the amount of organic nitrogen was very small the two processes yielded accordant results as in the following cases :-Organic nitrogen in 100,000 parts Organic nitrogen in 100,000 parts of water. of water. By permanganate By combustion. By permanganate By combustion. process. process. *001 .... -001 -004 .... -004 0001 . . .. *OOl 0003 .. . . -004 904 ... . -004 *002 .... *001 *010 .... 0009 -003 . ... -004 -012 .... -012 *002 .... *oo1 -001 .... *001 *002 .... *002 In a few other cases however the amount of organic nitrogen obtained by the permanganate proceas wa8 higher than that yielded by combustion as for instance :-Organic nitrogen in 100,000 parts Organic nitrogen in 100,000 parts of water.of water. By permanganate By combustion. By permanganate By combustion. process. process. -010 .... 0007 -004 .... -000 0009 .... 0005 *002 .... ,000 *003 .... *OOO -003 ... -000 ON THE ANALYSIS OF POTABLE WATERS. 101 These last results are to some extent explained by the fact that distilled water purified by boiling with alkaline potassic permanganate for a long time after ammonia has ceased to be evolved always yields ammonia when again treated with alkaline potassic permanganate. Thus in four experiments made with such purified water the following quantities of ammonia per 100,000 parts of water were obtained :-No.I. .... -002 part I No. 111. .... *002part No. 11. .... -001 , No. IV .... ,002 99 3. Estimation of h'itrogen in the fom of Nitrates and Nitrites. This determination can be made with very great accuracy by a modification of a process proposed twenty years ago by Walter Crum for the refraction of nitre.' It consists in agitating with mercury a concentrated solution of the nitrate or nitrite with a large excess of concentrated sulphuric acid when the whole of the nitrogen is evolved as nitric oxide. We find that for the succem of this process it is absolutely neces- sary that no chlorides should be present and also that the mixed liquids should be violently agitated witb mercury so as to break up the latter into minute globules.The following determinations show the accuracy of this pro- cess :-I. *02 grm. of nitre gave 75.48 C.C. nitric oxide at 49 mm. mercurial pressure and 16'94 C. 11. -01grm. of nitre dissolved in a saturated solution of sodic sulphate gave 75.48 C.C. of nitric oxide at 24.2 mm. mercurial pressure and 17O.8 C. Weight of nitrogen. Calculated. Found. No. I.. ,... -002772 -002897 No. I1 . . .. *001386 0001424 It was ascertained that uric acid hippuric acid urea and creatin when agitated with concentrated sulphuric acid and mercury gave no trace of gas. The following is the mode in which this process is applied to the estimation of nitrogen existing as nitrates and nitrites in potable waters.The solid residue from the half litre of water Phil. Mag. xxx 426. FRANKLAND AND ARMSTROKG used for determination No. 1 (estimation of total solid con- atituents)* is treated with a small quantity of distilled water a very excess of argentic sulphate is added to convert the chlorides present into sulphates and the filtered liquid is then concentrated by evaporation in a small beaker until it irJ mkhxd in bulk to two or three cubic centimetres. The liquid must now be transferred to a glass tube Fig. 2 and furnished at its upper extremity with a cup and stopcock pre- Viously filled with mercury at the mercurial trough the beaker being rinsed out once or twice with a very small volume of recently boiled distilled water and finally with pure and concentrated sulphuric acid in somewhat greater volume than that of the con- centrated solution and rinsings previously intro- duced into the tube.By a little dexterity it is easy to introduce successively the concentrated liquid rinsings and sulphiiric acid into the tube by means of the cup and stopcock without the admission of any trace of air. Should however air inadvertently gain admittance it is easily removed by depressing the tube in the mercury trough and then momen- tarily opening the stopcock. If this be done within a minute or two after the introduction of the sul- phuric acid no fear need be entertained of the loss of nitric oxide as the evolution of this gas does not begin until a minute or so after the violent agitation of the contents of the tube.The acid mixture being thus introduced the lower extre- mity of the tube is to be firmly closed by the thumb and the contents violently agitated by a simultaneous vertical and lateral movement in such a manner that there is always an unbroken column of mercury at least an inch long between the acid liquid and the thumb. From the description this manipu- lation may appear difficult but in practice it is extremely simple the acid liquid never coming in contact with the thumb. In about a minute from the commencement of the agitation a * If the water contain nitrites a separate half litre should be taken for this deter- mination otherwise there is a risk of loss of nitrogen during evaporation. The nitrites in tbis half litre of water must be transformed into nitrates by the cautious addition of potawic permanganate to the slightly acidified water before the evapo- ration is commenced.Immediately after the action of the permanganate the water must of coume be again rendered slightly alkaline. ON THE ANALYSIS OF POTABLE WATERS. Btrong pressure begins to be felt against the thumb of the operator and mercury spurts out in minute streams as nitric oxide gas ia evolved. The escape of the metal should be gently resisted so as to maintain a considerable excess of pressure inside the tube and thus prevent the possibility of air gaining access to the interior during the shaking. In from three to five minutes the reaction is completed and the nitric oxide may then be transferred to a suitable measuring apparatus where its volume is to be determined over mercury.As half a litre of water is used for the determination and as nitric oxide occupies exactly double thevolume of the nitrogen which it contains the volume of nitric oxide read off expresses the volume of nitrogen existing as nitrates and nitrites in one litre of the water. From the number so obtained the weight of nitrogen in these forms in 100,000 parts of water is easily calculated. 4. E.stimation of Ammonia. Unless the amount of ammonia obtained by diatillation alone or with sodic carbonate be considerable(above -01part in 100,000 parts of water) Hado w’s modification of Nessler’s process is all that could be desired for its accurate determination.But if a largey proportion than this be obtained the presence of urea may be suspected and it becomes necessary to make the Nes sler ammonia determination directly in the ori-ginal water without the intervention of distillation. For this purpose however the water should be colourless and free from calcic and magnesic carbonates. Any tint which is appreciable in a stratum 6 or 8 inches thick would obviously vitiate the result of a colour-test; whilst if calcic or magnesic carbonate be present the addition of the Nessler Rolution will infallibly produce turbidity ; moreover we find that the slightest opalescence in the water under these circumstances is abso- lutely incompatible with an accurate determination. Both these difficulties might be effectually removed by adding to the water first a few drops either of ferric chloride or aluminic chloride in solution and then a few drops of a solution of sodic carbonate so as to percipitate ferric hydrate or aluminic hydrate.The pre- cipitate completely decolorises the water and no turbidity is caused by the subsequent addition of the Nessler solution ;but unfortunately the precipitate carries down with it an amount FRANKLAND AND ARMSTRONO of ammonia which in the case of the ferric hydrate sometimes amounts to one-third of the total quantity present. Remem-bering the beautiful blue-green tint-the natural colour of absolutely pure water-which is presenied by a reservoir of water that has been softened by Clark's process we tried upon peaty water the effect of precipitating in it calcic carbo- nate and found that the decolorisation was as complete as could be desired arid that no appreciable amount of ammonia was carried down with the precipitate.The amount of calcic carbonate present in a coloured water is rarely sufficient to enable the operator to carry out this reactioii with sufficient rapidity and completeness; it is therefore best in all cases to add a few drops of a concentrated solution of calcic chloride to half a litre of the water. The subsequent addition of' a slight excess of' sodic carbonate then produces a copious precipitate of calcic carbonate which should be allowed to subside for half an hour before filtration. 100C.C. of the filtrate is a convenient quantity to take for the direct Nessler determination of am-monia.To this volume of the filtrate 1C.C. of the Nessler solution is added and the c'lour observed in the usual manner (see Miller on the Analysis of Potable Waters. Jour. Chem. SOC.,vol. xviii p. 125). By this direct process the ammonia in fresh urine can be readily estimated for this purpose 5 C.C. of the urine should be diluted with 95 C.C.of water free from ammonia. We have ascertained that known quantities of ammonia added in the form of ammonic chloride to urine can be determined with great accuracy. The colour observations of the Nessler determination are best made in narrow glass cylinders of such a dia- meter that 100 C.C.of the water to be tested form a stratum about seven inches deep.The depth of tint is t FIG 3 beat observed by placing these cylinders upon a sheet of white paper near a win- dow and looking at the SUT-face of the liquid obliquely; thus Fig. 3. ON THE ANALYSIS OF POTABLE WATERS. The nitrogen existing as nitrates nitrites and ammonia in potable waters is derived partly from the atmosphere and partly from the decomposition of nitrogenous organic matters previously existing either in the water or in the soil with which the water has been in contact. In view of the opinions now very generally entertained with regard to the propagation of certain forms of disease bymeans of spores or germs contained in excrementibious matters the search for nitrates nitrites and ammonia is second only in importance to that for actual sewage Contamination;because although these substances are in them- selves- innocuous unless present in excessive quantity yet when contained in a water in more than a certain propor-tion they betray previous contamination by sewage or by manured land.The nitrogenous organic matters contained in sewage or manure undergo slow oxidation and conver- sion into mineral compounds when mixed with water; their carbon is converted into carbonic anhydride and their hydrogen into water. These mineral products can no longer be identified in the aErated waters of a river spring or lake but the nitrogen is transformed into ammonia nitrous acid and nitric acid ; the two latter combine with the bases contained in most waters and together with the ammonia constitute a record of the sew- age or other analogous contamination from which the water has suffered.With certain corrections mentioned below the determination of the nitrogen contained in these mineral com- pounds proclaims the previous history of the water as regards its contact with decomposing nitrogenous organic matters. We propose to employ this determination for the expression of the previous sewage contamination of a water in terms of average filtered London sewage which if thus oxidised would yield a like amount of nitrogen in the form of ammonia nitrites and nitrates. For this purpose average filtered London sewage may be assumed to contain 10 parts of combined nitrogen in 100,000 parts as deduced from the numerow analyses of Hofmann and Witt and of Way and Odling.The number so obtained as the previous sewage contami- nation of a water requires however a correction since rain water itself contains combined nitrogen as ammonia nitrite of ammonia and nitrate of ammonia. The amount of these sub-stances present in the rain which falls at Rothampstead hag been determined by a series of monthly analyses made inde- FRANKLAND AND ARMSTROPU’O pendently on the one hand by Messrs. Lawes and Gilbert and on the other by Mr. Way and extending over two years. The results of these chemists give as the average amount of combined nitrogen 00985part in 100,000 parts of rain water. But as only a very small proportion of the rain water which supplies a river falls directly into the stream and as rain water is very rapidly deprived of its ammonia and to some extent also of its nitrites and nitrates by contact with vegetation this number as representing the amount of combined nitrogen conveyed into a river from uerial sources must.obviously be too high; indeed the experience gained in the examination of fifty samples of water collected near the source of streams proves this to be the case for the maximum amount of nitrogen as ammonia contained in any of these samples was only 0008part in 100,000 whilst the average amount of nitrogen in the form of nitrous and nitric acids observed by Messrs. Lawes Gil- bert and Way and this in thunder rain only was but *024 part in 100,000.It may therefore be safely assumed that hhe maximum amount of combined nitrogen derived by natural waters fiom aerial sources does not exceed 0024 + 0008= -032 part in 100,000; and we therefore propose to deduct this amount from the quantity of nitrogen present in a water in the form of ammonia and of nitrites and nitrates and to employ the remainder if any for the calculation of the previous sewage contamination on the basis that 10 parts of nitrogen correspond to 100,000 parts of such Contamination. If we represent the nitrogen existing in 100,000 parts of water as nitrates and nitrites by N and the nitrogen present as ammonia in the same quantity of water by N’ the previous sewage contamination of 100,000 parts of the water is denoted by the following expres- aion :-10,000 (N -C N -*032).Thus a water which contains in 100,000 parts *339 part of nitrogen as nitrates and nitrites and -001 part of ammonia has a previous sewage contamination of 3,080 parts; that is 100,000 parts of the water have been previously contaminated with sewage or manure matter equivalent to 3,080 parts of average filtered London sewage. Previous must not be con- founded with actual or present sewage contamination ; the latter is caused by unchanged or unoxidized sewage whilst the former ON THE ANALYSIS OF POTABLE WATERS. 1°7 denotes sewage completely resolved so far as its dead and un-organised organic constituents are concerned into peifectly or comparatively innocuous mineral compounds.But although this change has been effected at the time the sample of water was collected for analysis it by no means follows that it will be equally complete under future altered conditions as regards temperature exposure to air or vegetation and comparative volume of pure water. Previous sewage contamination must therefore to some extent be regarded as possible actual sewage contamination at some future time at the place where the sample was taken. There is also another aspect in which the previous sewage contamination of a water assumes a high degree of importance if the shell of an egg were broken and its contents beaten up with water and thrown into the Thames at Oxford the albumen would probably be entirely converted into mineral compounds before it reached Teddington; but no such destruction of the nitrogenous organic matter would ensue if the egg were carried down the stream unbroken for the same distance; the egg would even retain its vitality under circum- stances which would break up and destroy dead or unorganized organic matter.Now excrementitious matters certainly some- times if not always contain the germs or ova of organized beings; and as many of these can doubtless retain their vitality for a long time in water it follows that they can resist the oxidizing influences which destroy the excrementitious matters associated with them. Hence great previous sewage. contamination in a water means great risk of the presence of these germs which on account of their sparseness and minute size utterly elude the most delicate determinations of chemical analysis.These considerations respecting the import of the previous sewage contamination of a water lead us to regard from a sanitary point of view the accurate determination of the nitrogen in the form of nitrates nitrit,ee and ammonia as being next to the organic carbon and nitrogen determinations-by far the most important datum in water analysis. In illustration of our mode of expressing the results of water analyses the following table is subjoined. The degrees of hard- ness which we ernploy express the number of parts of calcic carbonate or its equivalent of other hardening salts in lOO,OoO parts of water ; they harmonize better than Clark’s with the 108 FRANKLAND AND ARMSTRONG ON POTABLE WATERS.decimal arrangement of the rest of the analytical results and if it be desired they are readily converted into Clark's degrees by multiplying by *7. The numbers opposite Thames water are the means of analyses of the water delivered by the Chelsea West Middlesex Southwark Grand Junction and Lambeth Companies on the 21st of January last. The New River East London and Kent Companied waters were collected about the same time. The lake watem were analised for the Roya€ Commission on Water Supply by Dr. Odling and one of US. The results yielded by the Caterham Company's water are interesting as an example of the great improvement effected in a chalk water by the application of Clark's process. Unfor-tunately this last sample was not taken under our supervision and we cannot therefore vouch for its authenticity.Ag 1 0, -9. fi 3 8 & a .+ 84 -gs *i Names of Waters. 3 3.g . n A & .-2 2 9 z.d *i $ a..E g;ka P a sn $8 g B 0 -4 -pc Thames water 86 delivered in London ..... 30.94 ,399 ,048 ,001 -395 3150 17 -3 River Lea water as delivered by New River Company ................................ 30-20 .115 *014 *001 -376 3300 20 '5 Ditto delivered by the East London Com-pany ............................................ 36*OO '147 ,024 ,001 -332 2760 22'8 Chalk water delivered by the Kent Com- pan y .............................................. 44-80 -064 '013 '001 -422 3770 26 -2 Ditto delivered by the South Essex Com-pany ................................................37-98 '164 000 -006 * 852 8205 21-6 Ditto supplied by Caterham Water works 12.80 '064 -007 '009 -014 0 7'0 Ditto supplied to W-orthing..................... 36.70 .162 -000 '000 *426 3940 23'8 Glasgow from Loch Katrine ..................... 3 -28 '256 -00s '002 -041 0 .3 Manchester from Derbyshire Hills ............ 6 80 -242 *026 .001 '028 0 2 '7 Lanester from Bleasdale Fells ............... x .54 *I57 -001 '001 '038 50 1 Preston from Longridge Fells .................. 14.70 '515 .040 -003 '044 0 6'7 Leicester waterworks supply .................. 23.70 -506 -020 '001 ,022 0 13-4 Bala Lake ........................................... 2 '79 -227 '001 '001 -002 0 '4 Ulswater Lake .......................................3 63 ,067 '000 '003 -007 0 1 '9 Thirlmere Lake ................................... 2-66 '194 .004 ,003 -008 0 '7 Haweewater Lake ................................. 3'56 -158 -004 004 -007 0 1 '3 Well water from Lepland near Preston, Lanwhire.......................................... 54 40 .325 -056 '003 '524 !4360 17 -6 Ditto from Ledbury .............................. 66.30 '145 -030 ,001 -606 15440 25-1 Ditto from Redhill ................................43 '60 .284 '021 '002 469 L4160 25.1

ISSN:0368-1769    

DOI:10.1039/JS8682100077

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

109 XII.-@n a Simple Apparatus fop determining the Gases incident to Water Andy&. By E. FRANKLAND, F.R.S. THEriianoroetric ga s-analysis apparatus described by Mr. Ward and myself,* enables the operator to make all the gaseous de- terminations connected with water-analysis with a rapidity delicacy and precision leaving little to be desired ; nevertheless as this instrument is also designed for the analysis of gaseous mixtures of greater complexity its construction is more elaborate than is necessary for the investigation of such gases as require to be dealt with in water-analysis. I have therefore devised a more simple and much less costly apparatus which permits of the rapid and accurate analysis of such gaseous mixtures as require to he submitted only to the action of absorbents,-a category which include3 all the gases appertaining to ordinary water- analysis.This apparatus is represented in fig. 1. A is a U-shaped glass tube 16 mm. internal diameter sup-ported in a perfectly perpendicular position by a convenient clamp. Its longer limb stands 1.1metre high its shorter limb 350 mm. measured to C. In the longer limb just above the bend is inserted a short glass tube b 1centimetre long and 2 mm. internal diameter ; it is attached to a doubly tubulated bottle B by a piece of strong caoutchouc tube 1; metre long and 2 mm. internal diameter. This bottle and tube serve for the supply of the apparatus with mercury. The shorter limb of the tube A is contracted at C and joined to a tube 220 mrn.long the bore of which does not exceed 13mm. To the upper extremity of C is joined the capillary tube d d' carrying a glass stopcock e. The upper extremity of d d' is bent horizontally and is carefully cemented into the steel cap and clamp-joint f the rrtructure of which is shown in section in figs. 2 and 3. gg represents a similar steel cap glass stopcock and capilla;ry tube connected with the small absorbing jar I 100 mm.high and 38 mm. internal diameter. The two steel caps permit of being joined gas-tight into a continuous c;lpillary tube by a Journ. Chem. SOC.,vi 197. VOL. XYI. I 110 FRANKLAND ON A SIMPLE APPARATUS FOR K DETERMINING THE GASES INCIDENT TO WATER-ANALYSIS. 111 small screw clamp fig. 3 as oiigiually devised by Regnault.The jar I stands fu-mly upon the shelf of a wooden mercury trough G. The construction of this tjrough is shown by figures IV V and VI. It is 265 mm. long 80 mm. broad and 90 mm. deep outside measure. The rim AA is 8 mm. broad and 15 mm. deep. The excavation B is 230 mm. long 26 mm. broad and 65 mm. deep. The shelf or bottom of the circular cavity on which the jar I,rests is sunk to a depth of 20 mm. below the top of the trough C or 35 rnm. below the top of the rim A (Fig. vi). At D in the excavation are two slight lateral indentations for the convenient transference of tubes containing specimens of gas from their capsules to the trough. Both limbs of the U-shaped tube are graduated in rnillimetres &om below upwards care being taken that when the tube is placed perpendicularly the zeros on both limbs are exactly at the same level.HH is a glass cylinder fixed upon the tube C by a perforated caoutchouc cork F slipped over the top of the tube. To facilitate the placing of the cork and cylinder in position the horizontal portion of the capillary tube d should be as short as possible. The cylinder H H serves to contain water intended to give without delay a definite temperature to gases which have to be measured in the limb C. To secure a Uniform temperature in this column of water an agitator hh 112 FRANKLASD OX A SINPLE APPARATCS FOR consisting of a copper wire flattened and bent into the form of a ring at its lower extremity is employed.Before this instru- ment is ready for use its shorter limb C requires calibration which is effected with great facility and accuracy in the follow- ing manner. The instrument is filledwith niercury by placing the reservoir B upon the stand K opening the clamp at b and the stopcock e after disconnecting 9 I at f. When mercury drips from the orifice at $ the cock e is shut. The jar I is iiow to be filled with mercury by applying suction to the orifice of its capillary tube until mercury issues from it and then closing the glass stopcock g. The tubes d and 9 must now be joined by the clitmp ,f and sufficient distilled water thrown up into I to fill completely the shorter limb of the U-shaped tube. The stopcocks e and g being opened and B again placed upon the table the clamp at b is to be unscrewed so as to allow mercury to flow from both limbs of the U-tube and thua to draw over the water from I into C.When the latter is so far filled with water as to depress tlie level of the mercury below the zero of the graduation the cock e is closed I is disconnected at f and removed and the mercury reservoir B again placed upon the stand K. The cock 4,is now cautiously opened and water allowed to drip from f until the convex surface of the mercury in C exactly marks the zero of the graduated scale. By greasing the face of the steel cap at f the water will flow down it in minute globules without wetting its surface. The tem-perature of the water in t4he cylinder HH being now noted the calibration may be commenced.A small and light glass flask the weight of which has been accurately ascertained is placed in such a position as to receive the drops of water when the latter are made to flow from f. The cock e is then cautiously opened and water allowed to drop into the flask until the mercury has risen in C a certain number of millinietres when the weight of water collected in the flask is ascertained. If the tube be of tolerably uniform bore it will be tsufficient to weigh the water at each rise of mercury through 100 millimetres bl1t the last reading of mercury in C should be taken at the highest point at which the calibre has remained unaltered in attaching the narrower tube and the next reading should in like manner he taken at the lowest point of the narrow tube where the bore has not been deformed by the glass-blower.It is obvious that between these two points no gaseous volume DETERMINING THE GASES INCIDENT TO WATER-ANALYSIS. 113 can be determined but the gas can always be either compressed into the narrower tube or expanded into the wide one by lengthening or shortening the column of mercury in the other limb of the apparatus. The operation of weighiug the water expelled between observed divisions of the scale is continued until the last drop is expelled from the capillary tube d. If the temperature of the water in H H be 4" C. the weight of water in grammea expelled from known lengths of the tube C ex-presses in cubic centimetres the respective capacities of those lengths of the tube.If the temperature of the water be not 4O C. the necessary correction must be made. It now only re- mains to calculate the volume-value of each millimetre in the different calibrated spaces and to prepare a table showing the volume of gas contained in the tube when the apex of the mercury stands at each individual millimetre from zero to d. The capillary depression of the mercury in the narrow tube must also be estimated by taking beveral readings in the two limbs when they are both freely open to the air. With a narrow tube of the diameter recommended above this capillarity will amount to about 3.4 millimetres which must therefore be de- ducted from the pressure in all determinations of gaseous volumes made within the narrow tube.The instrument is now ready for use. To preserve the mercury in the open limb of the U-shaped tube and in the reservoir B from dust and atmos- pheric impurity it is advisable to close their mouths with a loose plug of cotton wool. The following is the mode of conducting with this apparatus the analysis of-a gaseous mixture obtained in the combustion of a water residue. The gas is carefully transferred without loss to the jar I and thence for measurement into C,* where the apex of the mercury is brought to coincide with any millimetre mark. Let us suppose that the following observations are made Height of mercury in C ...................... 250 mm. Corresponding vol. of gas as per calibration table 19.200 C.C. Height of mercury in A ......................130.4 mm. Tempersture of water in H H ................ l6.4OC. Height of barometer ........................ 763.1 133131. * In the calibration the internal wulls of C are moietened with distilled water ; they are also ever aflerwsrds kept mobt so that the gaees when measured are always saturated with aqueoiis vnpour. FRAXKLAND ON A SIMPLE APPARATUS FOR From these data t,he pressure upon tjhe gas will be as fol- lows :-Height of barometer ........................ 763.1 mm. Deduct height of mercury column in A from height in C viz 250 -130.4 = 119*6 Plus tension of aqueous vapour at 16"-4C.= 13.9 -133.5 mm. Pressure on dry gas.. ........................ 629% mm- Hence we have 19.200 C.C. of dry gas at 16O.4 C.and 629.6 mm. pressure. Two or three drops of a Concentrated solution of caustic potash having been introduced by means of a small pipette into I the gas is now brought over into contact with it by placing the reservoir B upon the stand K and opening the clamp 6 and the stopcocks e and g. The absorption of carbonic anhydride is complete in about three minutes. The gas is again passed into C for measurement as before the difference between the two measurements giving obviously the volume of carbonic anhydride in the original mixture." The residual gas now consists of nitric oxide and nitrogen. Whilst it is still in the measuring tube C about an equal volume of oxygen is passed up into I and the gas being now brought over into the latter the nitric oxide is instantly transformed into nitric peroxide and absorbed by the excess of caustic potash present.But as excess of oxygen must be employed in this operation it is necessary to get rid of this gas before the volume of residual nitrogen can be determined. For this purpose two or three drops of a concentrated solution of pyrogallic acid are passed up into I. The oxygen is soon absorbed but its removal is much hastened by agitating the liquid so as to cause it again and again to moisten the surface of the glass receiver I. By dexterously causing the stand supporting the trough G to vibrate slightly this agitation of the enclosed liquid can be readily effected * If the combustion of the water residue be not pushed on too rapidly every trace of sulphurous anhydride will be absorbed by the metallic copper in front of the cupric oxide but it has occasionally happened in combustions very carelemly made that some sulphurous anhydride has eecaped absorption.If such a result be suspected it is only necessary to introduce two or three drops of a gatnrated solution of dipotassic diehromate into I before the original gas is transferred for measure-ment. DETERMINIXG THE GASES INCIDENT TO WATER-ANALYSIS. 115 without exposing the capillary tubes to the risk of fracture. The absorption of oxygen is known to be complcte when the dark-coloured liquid thrown by agitation upon the sides of the glass runs off again wit,hout leaving a dark blood-red stain. The operation is usually compIet,e at the expiration of five minutes.It now only remains to determine in C the volume of the residual nitrogen and t,he analysis is finished. The operationfi have furnished three uncorrected gaseous volumes viz. A volume of the three mixed gases; B volume of nitric oxide and nitrogen; and C volume of nitrogen. By the usual calculations these volumes may be reduced to 0" C. and 760 mm. pressure and then from the corrected volumefi A' B' and C' 80 obtained the quantities of carbonic anhydride and nitrogen may be deduced as follows :-A' -B' = vol. of carbonic anhydride. C' + B' = vol. of nitrogen. 2 From these corrected volumes of carbonic anhydride and nitrogen the weights of carbon and nitrogen can of course be readily calculated. This final and sole object of the analyst may however be obtained by a much shorter process; that is by dispensing altogether with the intermediate determination of the corrected volumes which is only necessary in the fore- going method of calculation in order tq enable the operator to arrive at the corrected volumes of carbonic anhydride and nitrogen.In fact if the original gaseous mixture be treated SO far as volume-weight is concerned as nitrogen the calcu- lations become greatly simplified. They depend upon the fol- lowing data :-1. The weights of carbon and nitrogen contained in equal volumes of carbonic anhydride and nitrogen gases measured at the same temperature and pressure are to each other as 6 14. 2. The weights of nitrogen contained in equal volumes of nitrogen and nitiic oxide are as 2 1.Now if we assume for the purposes of calculation that the gaseous mixture submitted to analysis consists entirely of nitrogen and that two successive portions of this nitrogen are removed from it by the action of reagents then if A be the FRANKLAKD ON A SIMPLE APPARATUS FOR weight of the total gas calculated as nitrogen B the weight after absorption of the first port,ion (CO,) and C the weight after the absorption of the second portion (N,O,); hrther if x and y represent respectively the weights of carbon and nitrogen actually contained in the gaseous mixture then the following simple equations express the values of x and y :-3 (A -B) X= 7 y=-C+B 2 By the use of the logarithmic table given below for the reduction of cubic centimetres of nitrogen to grms.for each -0012562 tenth of a degree centigrade by the formula (1 + *00367t)760 the labour of calculation is reduced to a minimum. An example will perhaps render the whole method of calculation here pro- posed more intelligible. For this purpose let us suppose the following values to have been obtained from readings such as those described at p. 113 :-Val. of original gas = A. 19.200 C.C. of dry gas at 16'94 C. and 629.6 mm. pressure. Vol. after first absorption (of CO,) = B. 3.342 C.C. of dry gag at 16"*7C. and 324.5 mm. pressure. Vol. after second absorption (of N,O,) = C. 1.631 C.C. of dry gas at 16"-9 C. and 298.4 mm. pressure. A= log. 19.200 = 1,28330 , 629'6 = 2,7990.6 log.*002562 for 16.4 C = -6,19286 (1 + *003671) 760 -2,27522 = e01885 gm. B= log. 3.342 = 0,52401 , 324.5 = 2.51121 log. (1 i-*00367t) 760 for 16O.7 C = -6,19241 -3,22763 = .001689grm. C= log. 1.631 = 0,21215 , 298.4 = 2,47480 .0012562 = -6,19211 log. (1 + *00367t) '160 for 16"-9C -4,87936 e= *0007575 ga. K 118 P'RMELANR ON A IFsMpr;bl APRAItATUI3 FUR Tablefor the Reduction of Cubic Centimetres of Nitrogen to Grams. *0012562 for each tenth af a degree from Oo to 3OO.C. Log. (1 .t '00367t)760 - . -. _ 7 t. d. - 8 -0 0.1 -0 *!a - o +a 0-4 -0 *5 -0 -6- 0 *7 0 -8-0.9- c- 6' 1 2 -6,mm665 507 808 649 491 793 633 475 777 617 459 P61 601 443 745 686 427 728 670 41 2 713 554 396 607 538 380 681 522 364 3 349 333 318 302 286 270 255 239 223 208 4 199 177 16I 146 f 30 114 098 08 3 067 051 5 035 020 004 "989 "973 #ssl '942 #926 "911 "895 6 -6,90879 864 848 839 817 801 786 770 755 739 7 723 708 692 6'76 661 645 629 614 598 583 8 667 552 556 521 605 490 474 459 443 428 9 413 307 382 388 381 335 820 304 289 274 10 a59 244 220 213 198 182 167 15.1 136 121 11 106 090 075 060 045 029 014 "999 "984 "969 12 -6,19953 938 923 907 892 87'7 862 846 831 816 18 800 785 770 765 740 734 709 694 679 661 11 648 633 6.18 603 688 573 658 543 528 613 15 497 482 467 452 437 422 407 392 377 362 Y6 a46 331 316 301 286 471 265 241 2!26 211 17 196 181 166 151 136 121 106 091 07i3 061 18 046 031 016 001 '986 "971 *956 #941 *926 *911 19 -6,18897 882 867 852 837 823 807 792 777 762 20 748 733 71 8 703 688 673 659 644 629 614 21 600 685 570 555 540 526 511 496 481 466 22 452 487 422 408 393 378 363 349 334 319 23 305 290 275 261 246 231 216 202 187 172 94 158 I48 128 114 a99 084 070 055 041 026 26 012 '997 *982 *968 '953 *938 *924 '909 "895 "88Q 26 -6,17866 851 837 822 808 793 779 764 250 735 27 721 706 692 677 663 648 634 619 605 590 28 676 561 547 532 518 503 489 475 460 446 29 432 417 403 388 374 360 345 331 316 302 - TabZe of the Elasticity of Apteow Vapour for each &th degree centigrade from 0' to 30' (&gnat&).dZ& .B g & 52b -blk n% d% .r( qg -:.Iremp remp 23g 2l co .CI Q) BZZ gsa C. C. :za Q) .Ii BX% EZ5f% 3xx 4 .6 6 *6 9 .4 -6 13.2 18-3 4-6 6-7 9.5 -7 13 3 184 4.7 6 -7 9 -5 -8 13 -4 18-5 4-7 6 '8 96 '9 13 *5 18-6 4 -7 6 '8 9 *7 16*O 13 -6 ia 4 *8 6-9 9 -7 -1 13 *6 18-8 4.8 6.9 9 *8 -2 13-7 19 -0 4.8 7.0 9.9 *3 13 .a 19-1 4 .9 7'0 9-9 *4 13 .9 19 -2 4.9 7'0 10 -0 -5 14-0 19 3 4.9 7 '1 10*I *6 14 *1 19.4 5.0 7.1 10 -1 9 14 *2 19*6 6.0 7 '2 10 *2 -8 14 2 19.t 5-0 7'2 10*3 '9 14 -3 19 .8 5.1 73 10-3 I7 '0 14 *4 19 -9 5.1 7.3 10.4 *1 14-5 20-0 5 '2 7.4 10 -5 -2 14-6 20 *1 5.2 7'4 10-5 .3 143' 20-3 5.2 7.5 10 6 -4 14.8 20 .4 5.3 7.5 10-7 -5 14-9 20.5 5.3 7.6 10 a7 .6 15-0 20.6 5'3 7 '6 10-8 *7 15 *l 20 -8 5.4 7.7 1039 -8 15 -2 20 .9 5.4 7 -8 10 -9 '9 15 -3 21 0 5 '5 7'8 11 -0 18'0 15.4 2s -1 5 '5 7.9 11 -1 '1 15 5 21 -3 5.5 7 '9 11 *2 *2 15 -6 21 -4 5.6 8.0 11.2 -3 15 .7 21 .5 5 '6 8.0 11a3 -4 15 -7 21 *7 5.6 8 .1 11 -4 -5 15 .8 21 *8 5 -7 8 -1 11-5 '6 15.9 21.9 57 8 -2 11 -5 -7 16 0 22 -1 5.8 8 -2 11 -6 .8 16 -1 22.2 5.8 8*S 11 -7 *9 16 -2 22*3 5 -8 8 -3 11 *8 19-0 16 .3 22-5 5.9 8 .4 11.8 '1 16 4 22.6 5.9 8.5 11.9 -2 16 .6 22-7 6 -0 8 *5 12.o -3 16-7 22*9 6.0 8.6 12.1 -4 16 .8 23 .O 6-1 8.6 12*1 -5 16 *9 23.1 6-1 8.2 12-2 -6 17 0 23.3 6-1 8.7 12 .3 -7 17 -1 23 .4 6 '2 a .a 12 -4 '8 17 -2 23 -5 6.2 8-9 12-5 -9 17a3 23-7 6 *3 8 *9 12 *5 30 -0 17*4 23 *8 6 *3 9.0 12.6 01 17-5 24*o 6 .4 9'0 12.7 -2 17*6 24.1 6 -4 9.1 12 8 *3 17 -7 24.3 6-4 9 *3 12.9 '4 17.8 24 4 6.5 9.2 12 *9 24 .6 6.5 9-3 13-0 24 *7 6*6 9.3 13 -1 24*8 - 4P4 884 614 PO9 069 949 09P EPP T8P 918 008 982 04 I 991 191 PZO+ 600r P66 848 898 SP8 TB4 914 TO4 P89 699 999 48P ZZP LOP 682 PLZ 692 TP1 981 1TI 266 446 Z96 Zt8 128 ZT8 269 449 299 TP9 9z9 TI9 068 948 098 68Z P22 602 880* 8401 890* 986 186 906 884 894 894 089 919 669 94f 'IUP EPP ZZ8 908 16Z 491 T9I 98T z IO# 966 186 998 OP8 PZ8 669 889 499 z t9 9z9 019 888 498 I98 PZZ 8i)Z Z6T 6.0 8.0 4.0 0.08 8.6Z 9 62 P*6Z 8. 68 T. 6Z 6.82 8. 8Z - $04 199 9 TP ILZ 9zr 086 P88 489 OT9 Z68 PPZ 960 4T6 464 tf 9 96P 9TE f6 I €PO* 068 884 f89 08P 9tZ oz T 996 608 T99 P6P 988 941 -9.0 - - 069 tP9 ZOP 49Z ZTI 996 6'18 ZL9 9Z9 448 6Z2 180 686 284 ZF9 T8f 088 64I 8201 9L8 zz4 699 91P 09z 901 096 €64 919 84P 6TE 09T -9.0 7 - 949 289 488 ZPZ 460 T 96 PO8 499 019 898 91z 990 416 494 419 997 918 P91 ZTOr 098 404 899 668 PPZ 680 P86 444 oz9 Z9P W8 PPT -f.0 - - 199 819 r4s (358 280 986 064 ZP9 96P SP8 002 190 Z06 Z9t zo9 I9T 008 6P1 466 PP8 169 889 788 6ZZ P40 816 194 TO9 4PP 882 6ZT -8.0 - - 4P9 109 898 1TZ 890 226 9t4 8Z9 18P GI8 981 980 488 464 489 9SP 982 P€ I Z86 6Z8 949 zz9 896 €12 890 806 974 689 l€P z4z 8TT -z.0 - - 889 68P ?PI 66 T €90 406 094 819 99P 818 04I TZO 248 ZZL z49 TZP otz 6lT 196 v 18 299 409 898 861 8PO 488 084 849 9IP 992 460 1.0 - -819 08 PAP 62 6ZE 82 P8T 42 88026'2 92 268 92 9P4 PZ 869 €8 K9P zz 808 12 991 02 90016% 61 498 81 tot LI 499 91 90P 91 992 PI €0T06'Z &I 196 ZT 864 TT 9P9 01 169 6 &8& 8 281 4 4206S'Z 9 It8 9 r14 P 499 8 668 Z OfZ 1 T8088'Z ,O _I 0.0 '07 -

ISSN:0368-1769    

DOI:10.1039/JS8682100109

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

121 XIII.-Reduction of Carbonic Acid to Oxalic Acid. By Dr. E. DRECHSEL. (Communicated by H. KOLBE.) AFTERmany unsuccessful experiments made during several years in my laboratory with the view of directly reducing car- bonic to oxalic acid-by a method similar to that by which Dr. R. Schmitt and I had effected its tramformation seven years ago-my assistant Dr. Drechsel has at length BUC-ceeded in solving this problem by very simple means. When a mixture of pure sodium and recently ignited dry sand is heated in a flask imbedded in a small sand-bath to the temperature of boiling mercury while a rapid stream of dry carbonic acid gas is continually passed through the vessel the pulpy mixture of sodium and sand which has at first a silvery lustre soon turns red and in a few hours the whole becomes converted into a dark-coloured pulverulent mass.Towarda the end of the operation care must be taken not to heat the mixture too strongly as in that case the product is apt to decompose with a glimmering light. The cooled mam is spread out on shallow dishes to facilitate the oxidation of the free sodium then exhausted with water and supersaturated with acetic acid ; and the oxalic acid is pre-cipitated from the filtrate by chloride of calcium. The precipi- tate is usually brownish; but by dissolving it in hydrochloric acid and neutralizing the filtered solution with ammonia the salt is obtained in the form of a snow-white powder. 60 grms. of sodium thus treated yielded ti grrns. of calcic oxalate. Dr. D r e chs el has subsequently found that potassium-amal- gam containing 2 p.c. potassium heated in carbonic acid gas to the boiling point of mercury absorbs the carbonic acid abundantly and yields a somewhat considerable quantity of oxalic acid. VOL. XXI. L

ISSN:0368-1769    

DOI:10.1039/JS8682100121

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

122 PERICIN ON SOME NEW BENZYLIC DERIVATIVES XIV.-On some new Benzylic Derivatives of the Xalicyl Series. By W. H. PERKIN,F.R.S. ON comparing the formulae of benzoin and benzylic acid with those of the hydride of salicyl and salicylic acid it will be observed that there exists a difference of C,H in their com-positian :- C14H1202 Benzoin. -c,H6 = c7H602 Hydride of Salicyl. C14H1203 -'VH6 Benzylicacid. = c7H603 Salicylic acid. If therefore an equivalent of hydrogen could be replaced by benzyl (C,H7) in the salicylic aldehyde and acid isomers of these substances would be produced. The following is an account of aome experiments which I have made in this direction and now beg leave to lay before the Society :-Action of Chloride of Benzyl upon the Hydride of Sodium-salicyl.On beating a mixture of equivalent quantities of the hydride of sodium-salicyl (salicylite of sodium) and chloride of benzyl with several times its volume of alcohol for three or four hours to a temperature of 120' to 140°C. in a sealed tube chemical action takes place with formation of chloride of sodium. After filtering off this salt and separating the excess of alcohol by evaporation a thick oily substance is obtained which when distilled yields only a small quantity of product below 320"C. by far the larger quantity coming over as a thick yellow oil above the range of the mercurial thermometer. To purify this oily distillate it is first agitated with a solution of hydrate of potassium and then with a strong solution of bisul-phite of sodium with which it slowly combines.It should be kept in contact with this reagent for two or three days with fkequent agitation or stirring otherwise a quantity of product may remain uncombined and be lost. The cryetabe com- OF THE SUICYL SERIES. 123 pound thus obtained is collected upon a cloth filter drained well squeezed from the excess of bisulphite of sodium and afterwards dissolved in very cold water. Ether is then added to remove oily impurities and the clear aqueous solution rendered alkaline with carbonate of sodium. The new body which has been liberated by this reagent is taken up with ether and upon evaporating this ethereal solution is obtained as a colourless viscid oil remaining fluid for days if left undis- turbed but gradually solidifying if agitated.It may then be rendered perfectly pure by recrystallisation from alcohol. The following combustiona of thk aubstance were made :-I. -1832 of substance gave *5331of CO, and -0976 of H,O. 11. 02661of substance gave -7706 of CO, and -1425 of H,O. 111. -2437 of substance gave -7062 of CO, and *1253of H,O. These numbers give percentages agreeing with the formula- aa the following comparisons will show :-Theory. Experiment. / 5 I. 11. 111. C, .... 168 79.24 79-31 78-97 79-03 H, .... 12 5 66 5.91 5-95 5.7-1 0 ...... 32 1510 --212 100~00 This substance represents the hydride of salicyl with its phenolic hydrogen replaced by benzyl. I therefore propose to call it the hydride of benzyl-salicyl.Its formation may be ex-pressed thus :-CO,H CO,H Hydride of sodium-Hydride of benzyl-salicyl. SalicyL L2 124 PERKIN ON SOME NEW BENZYLIC DERIVATNES As anticipated the hydride of benzyl-salicyl is not identical with benzoin but only isomeric. It melts at 46OC. benzoin melting at 120' C. With concentrated sulphuric acid it forms a yellow solution benzoin under the same circumstances pro- ducing a crimson one. With alcoholic hydrate of potassium it yields-a yellow liquid benzoin giving a violet-coloured re- action. The hydride of benzyl-salicyl boils at a temperature above the range of the mercurial thermometer. When cold it pos-sesses a slightly aromatic odour somewhat similar to that of cloves ; but when heated its vapour is both irritating and suffb-cating.It is easily soluble in ether tetrachloride of carbon benzol and likewise in boiling alcohol from which it crystallises on cooling in splendid transparent flat oblique prisms. In boiling water it dissolves to a small extent the solution becoming turbid on cooling and after standing depositing a small quantity of the aldehyde in crystals. Bromine and also nitric acid attack the hydride of benzyl-salicyl but the derivatives appear to be uncrystallisable bodies. As already seen this body is an aldehyde and combines with bisulphitefi. These combinations however do not form very easily when pure hydride of benzyl-salicyl is employed apparently on account of its solid condition and the insolubility of the resulting compounds in solution of the bisulphites ; tliere-fore it is better to employ the crude oily aldehyde in their preparation.The compound with biaulphite of sodium when crystallised over sulphuric acid forms beautiful small micaceous crystals possessing a very burning taste. When heated in a sealed tube with acetic anhydride to a temperature of 150°C. for three or four hours the hydride of benzyl yields an oily product apparently a direct combination. This compound if heated with water to 15OoC. decomposes with formation of acetic acid. Actio~cof Chloride of Benxyl upon Gaultherate of Sodium. Chloride of benzyl acts easily upon the gaultherate of sodium when heated with alcohol in a sealed tube of looo C.; four or five houra' digestion being generally sdcient to complete the OF THE SALlCYL SERIES.reaction. The product on being filtered off fiom the chloride of sodium which has formed and evaporated on the water-bath to remove alcohol yields a rather viscid oil which when recti- fied gives a considerable quantity of distillate boiling above 320' C. ; this consists of crude benzyl-salicylate of methyl. To obtain the acid fiom this product it is decomposed by boiling with alcoholic hydra.te of potassium. On separating the alcohol by evaporation a crude pasty potassium salt is obtained floating upon the excess of hydrate of' potassium which remains as a clear fluid and may be poured away. The potassium salt is then dissolved in water and agitated with ether to remove oily impurities and the clear aqueous solution decomposed with hydrochloric acid ; this causes the new acid to separate as an oil which fiolidifies in the course of twenty- four hours.It is then crystallised from alcohol once or twice or preferably fiom tetrachloride of carbon. Two combustions of fhh acid gave the following numbers :-I. -2570 of substance gave -6929 of CO and *1234of H,O. 11. *2452of substance gave -6627 of CO and -1224 of H,O. These numbers give per centages agreeing with those required by the formula ad^ the following comparisons will show Theory. Experiment. / h--I. 11. C14...... 168 73.68 73.53 73.70 H12 .... 1% 5.2 7 5-33 5-54 0,...... 48 21-05 --228 100*00 This substance represents salicylic acid in which the phenolic hydrogen is replaced by benzyl.I therefore propose to call it ben.zybsaZicyZic mid. Ibformation may be expremed thus :- 126 PERKIN ON Son NEW BENZYLIC DERIVATIVES. Sodium-salicylate Benzyl-salicylate of methyl of methyl. (Qaultherate of sodium.) 11. Benzyl-aalic ylate Benzyl-salicylate of methyl. of potassium. + HC1 = KCI. Benzyl-salicylat e Benzyl-salicylic of potassium. acid. Benzyl-salicylic acid melts at 75" C. and on cooling forms a viscid oil which crystallises in a confused manner when rubbed with a glass rod. It is extremely soluble in boiling and easily 80 in cold alcohol; it crystallises fiom this solvent in minute plates. In one experiment a portion of crude acid disaolved in Islightly diluted alcohol at first deposited an oil but after standing for about twenty-four hours beautiful transparent plates of the pure acid an eighth of an inch in diameter filled the oily deposit and jutted out into the clear solution.If boiled with water this acid dissolves to a small extent and the solu- tion on dooling becomes milky and after standing some time deposits the acid in thin brilliant plates. Benzyl-salicylic acid is isomeric with benzylic acid but does not give its coloured reactions neither does it give the violet coloration of salicylic acid with persalts of iron. Benzyl-salicy late of Ammonium.-Benzyl-salicylic acid dissolves freely in ammonia and on boiling off the excess of alkali a clear solution is obtained but this if evaporated to dryness decom- poses with separation of the acid.Benxyl-salicylute of Silver.-This body is obtained by the addition of nitrate of silver to the ammonium Balt. It i~ OF THE SALICYL SERIES. thrown down as a pure white precipitate slightly soluble in water. As a small quantity of benzyl-salicylic acid is often carried down with this compound especially if the ammonium salt employed in its preparation has been boiled rather too much it should be washed with alcohol as well as with water. It must be dried in vacuo as it fuses if heated in the water oven. A combustion and silver determinations gave the following results :-I. 01979of substance gave *3643of CO and 00635 of H,O. 11. -2194 of substance gave -0694 of silver.111. -1597 of aubstance gave 00508of silver. These numbers give per centagea agreeing with the formula as the following comparisons will show :-Tbe0ry. Experiment. % ,-r. II. IlI? C, .... 168 50-15 50.20 c -H, .... 11 3.2 8 3.56 -Ag .... 108 32.23 -31.63 31-80 -c 0,...... 48 14.34 -.. -c 335 100*00 BenzyFsaZicylate of Lead.-The addition of acetate of lead to a solution of benzyl-aalicylate of ammonium causes the salt to form aa a white curdy precipitate. BenzyE-saZicylate of Mercury ie a white precipitate. Benzyl-salicylate of Copper is a pale apple-green precipitate.

ISSN:0368-1769    

DOI:10.1039/JS8682100122

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

RU$SELL ON GAS ANALYSIS. XV.-On Gas Analysis. By W. J. RUSSELL,Ph.D. THEmethod of gas analysis proposed by Bunsen has this great advantage that it requires no complicated apparatus; a gra- duated glass tube with the means of holding it vertically in a mercury trough may be said to be all the special apparatus required. The reagent is used in the solid state and introduced into the eudiometer. The defect of this method as is well known is that the absorptions take place but slowly and that after touching the eudiometer it takes a considerable length of time to regain the temperature of the surrounding air so that several hours are required to complete an analysis by this method. These defects are obviated in the methods proposed by Re gn ault and R eiset and by Fr anklan d.These chemists use the reagent in the liquid state and surround the eudiometer with water. To do this it was however found necessary to have a more complicated form of apparatus ; the liquid reagent was not allowed to enter the eudiometer but was introduced into another tube termed the laboratory tube and the gas had by suitable mean8 to be passed backwards and forwards between the eudio- meter and this laboratory tube. In the apparatus which Dr. Williamson and I described in 1864" this transference of the gas was effected by a simple means fi-ee from any chance of loss or leakage. In all these forms of apparatus it is then principalIy the necessity of baving a separate tube €or the reagent which has rendered them so much more complicated than the apparatus of Buns en.My object now is to describe a method by which the liquid reagent can be introduced and removed fiom the eudiometer itself thus avoiding the necessity of a laboratory tube and the sluggish action of solid reagents. In what follows the process is applied to the method of Williamson and Russell above alluded to. The mercury trough is made of gutta-percha the form of it * Chem. SOC.Journ. xvii 238. RUSSELL ON UAS ANALYSIS. is shown in figs. 1and 2. Fig. 1 is a horizontal section and fig. 2 a section through A B. The larger part of the trough irJ circular ; in this part stands the glass cylinder which contains FIG. 1. Fra. 2. the water and hi the centre are the eudiometer and pressure- tube.The form of the well in which these tubes are raised or lowered is represented at C ;it is 24 inches long l+broad. The well for the pressure-tube is 14 inches deep measured fiom the bottom of the trough and that for the eudiometer 19 inches. These are shown in fig. 2 E and F. The sides of the trough are 34 inches high. The smaller part of the trough which is without the glass cylinder is shown in fig. 1. Along the bottom of this part there is a channel fj inch wide which runs into the well. The depth of this channel starting from the end of the bottom of the trough at B gradually increases till it reaches the well where it is 13 inches deep G H fig. 2. The circular part of the trough is 39 inches in diameter and the total length through A B 69 inches.The above dimensions are all inside measurements. The thickness of the gutta-percha is half au inch. RUSSELL ON GAS ANALYSIS. The pressure tube now used k simply a straight piece of tubing of about the aame diameter a8 the eudiometer; this simple form is more convenient FIa. 3. than the one which mas formerly proposed. The paper Bcreeii between the apparatus and the light is conveniently replaced by a strip of tissue-paper gummed on to the glass cylinder. Figure 3 represents the whole appa- ratus. The glass cylinder is omitted for the sake of clearness also the eudiometer but the clamp intended to hold it is very evident. The liquid reagent is intro-duced into the eudiometer by means of a small syringe readily made from a piece of tubing about 8 or 9 inches long and +th inch internal diameter; one end of it is bent round so as to give it the form of a hook and drawn out ;into the other end a piston fits made from a piece of stout steel wire one end of it roughened or a screw turned on it and round this cotton wool is tightly wrapped till it just fite the tube.In order to have it measure of the quantity of liquid to be introduced into the eudio- meter it is convenient to make five or six marks on the straight end of the syringe with a file a quarter of an inch apart. When the liquid is to be introduced a dot with a piece of chalk is made on the piston and it is then pushed down till this chalk dot reaches the mark corresponding to the volume of liquid to be injected.We now come to the question of how this liquid reagent is to be withdrawn without altering the bulk of gas in the RUSSELL ON GAS ANALYSIS. eudiometer. Several different processes were tried but it is only necessary to describe the one found to answer by far the best. The problem evidently was to find some body which could absorb a certain amount of liquid but which would not carry air with it when introduced into the eudiometer or abstract gas on being withdrawn. The following experiments will show that wet cotton wool has the required properties. It is used in this way a piece of steel wire size No. 9 or 10 has one end bent into a loop and gome cotton wool is twisted tightly round it ; this mass of wool should be about 8 inch in diameter and 2 inch long.It is placed in a basin of water and thoroughly kneaded and squeezed for some time ; this treatment wets the whole mas8 of cotton wool and expels all the air adhering to it. This ball of cotton on the steel wire might of course be in- troduced into the eudiometer in the same way as Bunsen introduces his solid reagents; but it is found much more con-venient especially where it is important to have the mercury trough small to use what may be called a guide tube. This is merely a piece of fine glass tubing about 8 or 9 inches in length and bent into a curve at one end. The steel wire is introduced at this elid and pulled through till the cotton ball prevents its going further.Held in this way the ball is easily introduced or withdrawn from the eudiometer and that with- out dipping the fingers into the mercury. The ball being now held in the tube and thoroughly saturated with water is lifted from t,he basin and plunged below the surface of the mercury in the trough; it is then squeezed between the finger and thumb so as to expel a qonsiderable portion of water but still to leave it very wet. The guide tube is now introduced into the canal D fig. 1 of the trough and pushed down it so that the curved end with the ball comes within the tall glass cylin- der. In order to introduce the ball into the eudiometer the eudiometer is raised so that the open end of it is a little above the well in the trough a mark on the cylinder or on one of the supports will indicate the point to which the top of the eudiometer must be raised in order that the open end of it may be in the right position.Thus raised it is easy to hook the cotton ball project- ing from the guide-tube into the eudiometer ; then by pushing the wire the ball rises in the tube and is completely under control. It is well to apply a very little grease to the wire and if by any RUSSELL ON GAS ANALYSIS. chance it should become bent it must of course be discarded. To withdraw the reagent the eudiometer is raised and the cotton ball introduced as above described; it is pushed up till the top of it comes in contact with the reagent which is then quickly taken up by the cotton-wool and the miniscus left free from liquid.The ball is now pulled below the surfhe of the mercury; thils should be done with a jerk to prevent any gas adhering to it and withdrawn from the trough. The following consecutive experiments will show with what great accizracy this removal of the reagent may be effected by this process. The experiments were the first ones made to test the accuracy of the process. Instead of using a reagent water was introduced into the eudiometer by means of course of the syringe. The numbers indicate the volume of liquid used ;for instance six of water means that six quarter-inches of this syringe-tube +-inch diameter full of water were used in that experiment. Volume of air taken ..................... 300.4 Do. after withdrawing 6 of water ....300.5 Do. do. .... 300.5 DO. do. .... 300.5 Do. do. .... 300.5 Volume of air taken ...................... 300.0 Do. after withdrawing 6 of water another ball of cotton used ...... 300.0 Volume of air taken ...................... 366.3 Do. after withdrawing 6 of water .... 366.2 Do. do. .... 366.2 D0. do. (another and larger ball used) .... 366.1 D0. do. .... 366.1 Volume of air taken ....................... 88.7 Do. after withdrawing 6 of water .... 88.7 DO. do. .... 88.6 Do. do. .... 88.7 In this cme the gw was very much expanded there being RUS3ELL ON GAS ANALYSIS. when the bulk wap introduced. a column of mercury 18 inches high in the eudiometer . Volume of air taken ......................89.0 Do. after withdrawing 6 of water ...... 89.0 Do. after adding carbonic acid .......... 111.3 Do. after withdrawing 6 of potash ...... 89.0 DO. do. ...... 89.0 Volume of air taken ...................... 169.5 After withdrawing 7 of potash .............. Do. 6 of potash .............. DO. do............... Volume of air taken ...................... Do. after adding carbonic acid .......... After withdrawing 6 of potash.............. Do. do. .............. Volume after adding carbonic acid .......... After withdrawing 8 of potash .............. Volume after adding carbonic acid .......... After withdrawing 6 of potash.............. Do. do............... Volume after adding carbonic add .......... After withdrawing 10 of potash ............Do. 6 of potmh ............ Volume of air taken ...................... After withdrawing 3 of water .............. Do. do............... After withdrawing 2 of water .............. Do. do............... Do. 6 of water ............. Do. 2 of potash .............. Do. do............... Do. 6 of potash new ball used. ball up twice .......... Do. 6 of potash. Tepeated 3 times ................ 16'3.4 169.4 169.4 116.5 141.0 116% 116.4 167-0 116.4 210.0 148.0 11 6.4 225-0 139-0 116-3to 4 133.6 133.6 133.6 133.6 133.6 133.6 133.6 133.6 133.5 133.5 These experiments will be sufficient to show the general RUSSELL ON GAS ANALYSIS.accuracy of the process . Connected with the use of each of the ordinary reagents. there are some points it will be necessary to allude to. Absorption of Carbonic Acid.-The following experiments were made with air and known quantities of carbonic acid. to test the accuracy of the method :-Volume of air taken ...................... 189.0 After adding carbonic acid ............... 204.0 The gas absorbed. no reading made . After adding carbonic acid ................ 209*0 The gas absorbed. no reading made . After adding carbonic acid ................ 211.0 The gas absorbed ........................ 189.0 To absorb the gas in each case. six of potash were used. and allowed to remain in the eudiometer ten minutes . The potash solution was made from a saturated solution of caustic potash.diluted with twice its bulk of water . In the following experi- ments when the term potash is used. a solution of this st,rength is always meant :- Volume of air taken ...................... 278.4 ARer adding carbonic acid ............... 305.6 After 5 potash. up 15minutes .............. 278.3 Volume of air taken ...................... 340.9 After adding carbonic acid ................ 380.1 After 5 potash up 18 hours ................ 340.8 Volume of air taken ..................... 325.2 After adding carbonic acid ................ 338.3 After 5 potash. up 20 minutes .............. 325.1 Volume of air taken ...................... 326.3 After adding carbonic acid ................335.9 After 5 potash. up 20 minutes .............. 326.3 Volume of air taken ...................... 333.1 After adding carbonic acid ................ 345.0 After 5 potash up 20 minutes .............. 333.0 RUSSELL ON QAS ANALYSIS. Volume of air taken ...................... 327.1 After adding carbonic acid ................ 332.13 After 5 potash up 20 minutes .............. 327.0 The foregoing seiies of experiments are sufficient to show the accuracy with which this absorption of carbonic acid may be made. Although these results are satisfactory it was still a question whether under other circumstances the intro- duction of a certain amount of caustic potash might not alter the tension of the gas even in a wet eudiometer.To investi-gate this point moist air free from carbonic acid was first taken and potash brought in contact with it in the eudio- meter. Volume of air taken ...................... 341.8 After 5 potash up 20 minutes .............. 341.5 Volume of air taken ...................... 333.7 After 5 potash.. .......................... 333.7 Volume of air taken ...................... 327.9 After 5 potash up 15 minutes .............. 327.5 Volume of air taken ...................... 140.7 After 8 potash up all night ................ 140.0 Volume of air taken ...................... 191.2 After 5 potash up 5 minutes .............. 190.9 Do. 10 minutes .............. 190.8 Do. do. .............. 190.8 DO. do. .............. 190.7 Volume of air taken ......................169.9 After 5 potash up 10 minutes .............. 169.6 Do. do. .............. 169-6 The introduction of five of potash into a gas containing no carbonic acid causes a slight diminution of tension. By means of the syringe a bubble of water is easily introduced into the gas after the removal of the potash; but unless comparatively large quantities of water are used which of course would be inadmissible the original tension of the gas is not generally restored. 136 RUSSELL ON GAS ANALYSIS. Volume of air taken ...................... 204'7 After 5 potash. up 10 minutes .............. 204.5 .. 5 water. up 10 minutes .............. 204.5 .. 5 water. up all night ................ 204.45 ..5 potash. up 10 minutes .............. 204.4 .. 5 strong potash. up 10 minutes ........ 204.1 .. 5 water ............................ 204.45 The strong potash consisted of 2 parts of a saturated solution of caustic potarJh 1 part of water. Volume of air taken ...................... 318.4 Afier 3 potash. up 15 minutes. tube lowered 3 times ............................... 318.0 After 4 water. up 10 minutes. tube lowered 2 times ................................ 318.3 After 3 potash. up 10 minutes. tube lowered 3 times ................................ 318.0 After 3 water. up 10 minutes. tube lowered 3 times ............................... 318.3 After 3 potash. up 10 minutes. tube lowered 3 times ...............................318.0 Volume of air taken ...................... 315.2 After 6 potash. up 15 minutes .............. 314.8 .. 5 .. 10 minutes ............. 314.8 .. 5 .. all night ................ 314.6 .. 5 water up .......................... 314.7 Y? .. .......................... 314% Volume of air taken ...................... 121.9 After 6 potash. up all night ................ 121.4 .. water up ............................ 121.6 These experiments. and many others which have been made. show that. under certain circumstances. a very appreciable error may be made by the potash altering the tension of the aqueous vapour. It appears. however. that when a considerable quan- tity of carbonic acid is present. some twenty or more milli-metres of that gas.this diminution of tension does:not take place. as i~3seen in some of the analyses already quoted. and. in some RUSSELL ON GAS ANALYSIS. cases with even a much less quantity of earbonic acid it does not take place. This source of error is however easily avoided by using a much weaker solution of potash; one made by diluting 1 pt. of a saturated solution of potash with 5 pts. of water answers very well. In the following experiments a solution of this degree of dilution is termed ‘‘weak potash.” If 5 pts. of this solution are introduced into a moist gas even entirely free from carbonic acid it will produce no appreciable alteration in the tension of the aqueous vapour as shown by the following ex- periments :-Volume of air. tahn .....................321.6 After 5 weak potash up 15 minutes ........ 321.6 Volume of air taken ....................... 338.1 After 5 weak potash up 45 minutes ....... 337.95 Volume of air taken ...................... 281.1 After 5 weak potash up 15 minutes ........ 282.1 In the following experiments carbonic acid was present. Volume of air taken ...................... 188.6 After adding carbonic acid.. ................ 190.0 , 2 weak potash up 20 minutes,. ........ 188.5 Volume of air taken ...................... 243.6 Do. after adding carbonic acid .......... 313.4 After 5 potash up 2 hours.. ................. 244.9 , 3 weak potash up 10 minutes.. ......... 243.5 , adding carbonic acid.. ................ 255.3 , 5 potash up 30 minutes ..............243.5 Volume of air taken ...................... 339.4 After adding carbonic acid. ................. 340.3 , 5 weak pota.sh up 15 minutes.. ........ 339.4 The absorption of carbonic acid by this process is then per-fectly simple and accurate; the oiily precaution to be taken is when only testing for the presence of the gas or absorbing very small quantities to use a dilute solution in other cases it is VOL. XXI. M RUSSELL ON QAS ANALYSIS. better to we a stronger one. The 5 of potash used in most of the foregoing experiments would absorb about 80 millimetres of carbonic add measured in a eudiometer about Hth inch in diameter and under a pressure of about 8 inches less than the ordinary barometric pressure.The weak potash contains only half the amount of potash which the stronger does. In some cases it is convenient to use even a still stronger solution-one containing 2 pts. of saturated potash to one of water. In all the absorptions of carbonic acid after introducing the caustic potash the eudiometer was lowered so as to wet with the solution a considerable portion of the tube and thus expose a large surface of the solution to the gas. This raising and lowering of the tube should be repeated; the absorption then takes place very rapidly. The time required to absorb the whole of the carbonic acid depends on the absolute and relative amount of the gns present and on the amount of potash solu- tion used. Generally about ten minutes is ample time.A rapid observation of the height of the mercury in the eudio- meter with the telescope will ahow whether any further ab- sorption is taking place. In the experiments when a diminution of tenlsion was pro-dnced by introducing potash into moist air a few drops of water were of course as in other cases first placed at the closed end of the eudiometer ;it was then filled with mercury and the air introduced the upper part of the tube thus becoming moistened all over and covered with visible drops of water ; in raising and lowering the tube the potash did not come within three or four inches of the top of the eudiometer so that although the tension of the aqueous vapour diminished there was still an abundance of water in the top of the tube so much as to be perfectly visible to the naked eye.It ap-pears then that in the upper part of the tube the original full tension of the aqueous vapour must have continued ;but below where some of the potash would be still sticking to the sides of the tube a diminution of tension was produced. A con-tinual evaporating of the water at the top and diluting of the potash below is going on but is not sufficient iu amount and rapidity to restore the original tension. Volume of air taken .... . . ... . ... . ..... . . . 330.7 After 5 potash up 10 minutes ..........,... 339.2 RUSSELL ON GAS ANALYSIS. When this reading was made the upper part of the eudiometer in fact nearly 5 inches from the top was covered with rJmall drops of water. The eudiometer was allowed to stand for 50 hours and the bulk of gas then read off it was 330.3.The top of the tube now looked perfectly dry.* Absorption of Oxygen.-This gas is absorbed by finst putting up into the eudiometer some potash and then with another syringe adding a little of a strong solution of pyrogallic acid to it. The quantities must of course depend on the amount of oxygen present. To test the process some air waa taken which containing so large an amount of oxygen would show any defects in the process. The following are the results obtained the quantities of the reagents used and the time they were allowed to act :-Corrected v01. Volume of air taken .......... 130.3 132.15 4 potash and 3 pyrogallic acid solution left up half hour ....103.5 104.46 2 potash and 2 pyrogallic acid solution left up half hour .... 103.5 Oxygen .,. . .. 20.953 per cent. Volume of air taken ......... 129.0 130.80 4 potash and 3 pyrogallic acid solution left up 20 minutes .. 103.3 104.26 3 potash and 2 pyrogallic acid aolution left up 15 minutes .. 102.6 103.53 Oxygen ........ 20.848 per cent Volume of air taken .......... 135.4 137.44 * The following experiment may also be cited as illustrating in a different way the absorption of aqueous vaponr. Threc portions of cnlcic chloride of about 10grma. each were taken ; one portion placed in a watch-glass another at the bottom of a glass cylinder 43-in. high and l#-in. in diameter and the third at the bottom of a cylinder 8;-in. high and l*in.in diameter; all theae glasses were placed close together on a shelf and left uncovered. On the next day the calcic chloride in the watch-glass was liquid; on the thirteenth day that in the shorter cylinder was liquid but it was not till the twentyfirat day that the chloride in the tallest cylinder wae liquid. M:! RUSSELL ON GAS ANALYSIS. Corrected vol. 4 potash and 3 pyrogallic acid solution left up half hour .... 107.7 108-79 Oxygen ........ 20-845 per cent. Absorption of Olejant Gas.-This gas is absorbed in the ordi- nary way with fuming sulphuric acid ; the only question was whether the acid should be introduced by means of a syringe or bya coke ball. The latter process is perhaps the best. The coke ball is fastened on a platinum wire and with the guide tube is very easily introduced into the eudiometer.The same ball may be used a great number of times. After the ball haB absorbed all the olefiant gas (it will generally take from haIf-an- hour to an hour to do this) a little water is introduced into the tube this withdrawn and then some dilute potash. The following are some instances of the absorption of car-bonic acid and olefiant gas in coal gas; the specimens oDerated I on were from the same works but not collected at the same time :-Corrected vol. Volume of gas taken .......... 217.8 229.11 After 5 weak potash .......... 214.7 225.82 After coke ball and absorption of acid fumes.. ................ 202.8 213.21 CO .... 1.950 p. c. C,H, . .5.504 , Volume of gas taken .......... 214.2 225.29 After 5 weak potash .......... 211.0 22 1-90 After coke ball and absoi.ption of acid fumes.. ................ 200.0 210-25 CO ... 1.505 p. c. CnHzn .. 5.171 99 In the following analysis no coke ball was used the acid being introduced by means of a syringe. The carbonic acid and acid fumes were absorbed at one operation. Corrected vol. Volume of gas taken .......... 166.4 174.79 STENHOUSE ON CHLORANIL. Corrected voL 4 fuming acid up. afterwards phos- phate of Boda. and then potash 155.4 163.34 CO. and CnH.. 6.551 p... The following is a complete analysis of a specimen of coal gas :-Volume of gas taken ............ After 5 weak potash .............. ..coke ball .................. .. 4 potash and 2 pyrogallic acid Volume of combustible gases taken After adding oxygen ............ .. explosion ............... .. potash .................... .. adding hydrogen........... .. explosion ................. 229.11 225-82 213.21 202.5 51.82 171.04 85-38 60.08 411-40 249.47 This gives the following percentage composition for the gas :-Carbonic acid ................... Olefiant gas. &c ................... Oxygen ........................ Hydrogen ...................... Light carburetted hydrogen ...... Carbonic oxide .................. Nitrogen ........................ 1.950 5504 0.139 45.847 40.948 4.167 1.445 100.000

ISSN:0368-1769    

DOI:10.1039/JS8682100128

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

STENHOUSE ON C€?LORANIL. XVL-On Chloranil. No. I. By JOHNSTENHOUSE, LL.D. F.R.S. &C Preparation of ChZoranil. THEbest process for the preparation of chloranil hitherto pub-lished is that of Dr. Hofmann,* which consists in digesting * Ann. Pharm. lij 65. STENHOUSE ON CHLORAWIL a boiling saturated aqueous solution of phenol with a mixture of chlorate of potassium and hydrochloric acid. When chlorate of potassium and phenol are dissolved in hot waster and a sufficient quantity of' hydrochloric acid is added the solution becomes dark-coloured and turbid and after the lapse of a few minutes a strong reaction sets in with evolution of pungent vapours and eeparation of yellow crystalline scales which are however contamiiiated with a considerable quantity of a dark red oil.This mam when cold is freed as much as possible from water and extracted with hot spirit to get rid of the red oil. The pale yellow crystalline scales thus obtained consist of terchlorquinone mixed with fiom 20-50 per cent. of chloranil. By repeated treatment with boiling spirit the ter- chlorquinone as it is rather soluble in that menstruum may be removed and the chloranil obtained in a state of tolerable purity. The terchlorquinone thus extracted contains chloranil and although I have operated upon very large quantities I have never been able to obtain it free from adhering traces of the latter substance either by sublimation or by crystallisation fiom spirit or benzol; but as will presently be seen I have been able to obtain it pure indirectly.After making numerous experiments to determine the most advantageous proportions foe preparing chloranil according to the above method the following was found to be the best pro- cess :-3 parts chlornte of potassium were dissolved in 70 parts boiling water and 1 part phenol added. This mixture when poured into an earthenware vessel capable of containing about twice the volume of the liquid had a temperature of 90' C. ; 14 parts hydrochloric acid of sp. gr. 1.16 were then added at once and the whole was well agitated. In a few minutes the clear brownish-red liquid became opaque and very hot chlorine and a pungent vapour of very disagreeable and persistent odour were given off with violent effervescence and chloranil was deposited in yellow scales contaminated with the before-men- tioned red oil.On extracting this with a large quantity of boiling spirit chloranil is left undiBsohed but even when a great excess of chlorate of potassium and hydrdchloric acid was used I never suoceeded in obtaining more than 40 parts of chloranil f%om I00 of phenol. It seemed to me however that the terchlor- quinone and chlorinated oil holding 8s they do an intermediate STENHOUSE ON CHLORANLL. position between phenol and chloranil might be converted into the latter by proper treatment; and finding from a preliminary expeiiment thak chloride of iodine rapidly converted both of them into chloraiiil I resolved to utilize this reaction for the preparation of that substance.The above-mentioned mixture of crude chloranil terchlor- quinone and red oil obtained from phenol by chlorate of potassium and hydrochloric acid after standing from 12-24 hours was collected on a cloth filter fieed as much as possible &om water by pressure and introduced into a flask with an equal weight of water and about half its weight of iodine. This flask was fiirnished with two tubes one serving as a con- denser and the other for passing a current of chlorine into the mixture whilst heat was applied by means of a paraffin bath. The current of chlorine which was at first very rapidly ab- sorbed was so adjusted that there was always a slight excess of iodine. This was readily ascertained by the violet colour of the vapour in the flask.After 10 or 12 hours when the ab- sorption of chlorine became very slow the digestion was shopped a bent tube fitted to the flask in place of the condenser and the chloride of iodine solution distilled off as far as possible a gentle current of chlorine being passed though the apparatus during the latter part of the operation. When cold the residue was boiled up with water and the yellow scales which were now almost free from oily matter well washed with cold water pressed and extracted several times wit>h small quantities of spirit. It now presented the appearance of brownish-yellow 8calcs and was tolerably pure chloranil considerably more than equal in weight (1.25) to the phenol originally employed. Attempts were also made to convert both the red oil and terchlorquinone into chloranil by treating them with chlorate of potassium and hydrochloiic acid also by digestion with penta- chloride of antimony and sesquichloride of iron in a current of chlorine but these methods gave very unsatisfactory results as did also passing chlorine through the solution of the tjubstances in tetrachloride of carbon either with water or without.In order to purify the chloranil obtained by the above method one part of it was dissolved in 20 parts of hot benzol filtered and the excess of benzol distilled off until scales began to deposit in the hot solution; the distillation was then discon- tinued and when cold the chloranil was collected on a cloth STENHOUSE ON CHLORANIL.and strongly pressed. One or two cry stallisations thus rendered it perfectly pure. It is necessary that the benzol should be purified by sulphuric acid," as otherwise the chloranil becomes contaminated with a large quantity of black resinous substance. This is a better method of treating the chloranil than crystalli- sation .from boiling alcohol as it is but slightly soluble in that liquid. Analyses of Chloranil. I. -245 grm. substance gave -571 grm argentic chloride. 11. -294grm. substance gave ,686 grm. argentic chloride. I. II. c C1 0 = = = 72 142 32 - 29.26 57-73 13-01 -57.65 - -57-73 - 246 100.00 I. was obtained by the action of chloride of iodine on the dark red oil and 11. .from terchlorquinone containing traces of chloranil by the same method.When phenol previously combined with sulphuric acid (equal measures) was dissolved in a large quantity of hot water and submitted to the action of chlorate of potassium and hydro- chloric acid no red oil was formed but the amount of chloranil and terchlorquinone was less than that obtained by acting on a aolution of pure phenol in water. Prepayation OJ Chloranil from Picric Acid. On dissolving equal parts of picric acid and chIorate of potassium in 30 parts boiling water a considerable quantity of the difficultly soluble picrate of potassium was formed and the addition of hydrochloric acid (7 partsj caused a violent re-action ; chloropicrin distilled over and chloranil remained in the retort equal to about one-twelfth the weight of the picric acid originally employed being only 12 per cent.of the theoretical * Proc. Roy. SOC.xiv 351. STENHOUSE OK CHLORANIL. quantity. This therefore is not by any means an advanta-geous process. Chloranil is but dightly soluble in bisulphide of ca.rbon tetrachloride of carbon ether chloroform or petro- leum oil. Cldorhydranil. Stideler in 1849," prepared chlorhydranil by digesting chloranil with boiling aqueous sulphurous acid until it ex-hibited po further change of colour. He believed that chlorhy- dranil was the only product formed by this reaction. This however was a mistake as I found on carefully repeating his process that 100 parts chloranil gave only 70 chlorhydranil instead of 101 and that the crude chlorhydranil thus obtained had a brown colour which is very difficult to remove.A much better method which gave the theoretical quantity of colour- less chlorhydranil at one operat.ion consisted in digesting finely powdered pure chloranil with moderately strong hydriodic acid and about one-tenth its weight of ordinary phosphorus for 30-40 minutes washing well with cold water and crystal- hing fiom boiling alcohol (5 parts). The product thug obtained was quite colourless and of a high lustre. It wag however contaminated with traces of phosphorus which soon became oxidised on exposure to the air and the resulting phosphorous acid was removed by washing with cold water. This impurity can also be got rid of by boiling with aqueous cupric sulphate and recrystallising.The substitution of amor- phous for ordinary phosphorus in the above process is not advisable as the crystals are then slightly coloured. Chlorhydranil C,Cl,O,H, is almost insoluble in bisulphide of carbon tetrachloride of carbon and benzol but very soluble in ether. It dissolves in carbonate of sodium solution with a beau- tiful green colour which rapidly changes into brown and at the same time a few green needles are deposited. By heating with strong nitric acid or better still a mixture of nitric and sulphuric acids it is easily reconverted into chloranil. Tbis method may be employed for the purification of crude chloranil instead of crystallisation from benzol. 8 Ann. Ch. Pharm. Ixix 327. STENHOUSE ON CHLORANIL. Action of Sulphurous Acid on Chloranil.Terchlorhydropuinone. Although chlorhydranil was the principal product obtained by passing a current of sulphurous acid through boiling water holding chloranil in suspension yet as I have mentioned above a'bout 30 per cent. of it was converted into ot'her compounds which remained in the aqueous solution contaminated with sulphuric and hydrochloric acids. This solution was neutralized with carbonate of lead (pre- viously made into a cream with water) and filtered. The aulphuric acid was thus removed as sulphate of lead whilst plumbic chloride remained in solution together with an organic lead salt. The lead was precipitated fi-om this solution by sulphuretted hydrogen the clear liquor evaporated to dryness on the water-bath and the residue consisting of an organic acid submitted to sublimatiou which was best effected by placing it in a covered beaker heated by means of a paraffin bath to 12Oo-13O0 C.white lustrous crystals were thus obtained. These were purified &om adhering hydrochloric acid by one or two resublimations or by recrystallising fkom bisulphide of carbon and then subliming them. The residue in the sublirn- ing vessel contained free sulphuric acid. The crystals are moderately soluble in hot water alcohol bisulphide and tetra- chloride of carbon and very soluble in ether. When subjected to analysis they gave the following results :-I. 0229 grm. substance gave -282grm. carbonic acid and -030 grm. water. 11. 976 grm. substance gave -341 grm.carbonic acid and *036 grm.water. 111. -180 grm. subatance gave 362 grm.argentic chloride. I. 11. 111. c = 72 33.72 33-59P.C. 33.70 P.C. -c1 = 106.5 49-89 -49-75p.c. €I,= 3 1-41 1.45 1.45 -0 = 32 14-98 ---213.5 100*00 It will be seen from thae results that the crystalline eubli- STENHOUSE ON C€WORANLL. mate had the composition of terchlorhydroquinone C6C13H,02. Its properties also agreed with those ascribed to that substance by Stadeler.* On adding nit,rate of silver to a warm solution of the crystals a white precipitate was produced containing 50 per cent. of silver. It appears however to be a mixture of a silver compound with terchlorquinone. Nitric acid acted strongly on terchlorbydroquinone red fumes were evolved chloropicrin distilled over and the solution on cooling deposited yellow crystals a further quantity being obtained on the addition of water.They were well washed wit.h cold water and crystallised from alcohol. As might be expected the analysis of the substance proved it to be ter- chlorquin on e C6C13 HO2. I. *281 grm. substance gave -349 grm. carbonic acid and water.0014grm. I. C,C1 = = 72 106.5 34-04 50.36 33.88 p.c. - H = 1 -47 *55 0 = 32 - 15.13 - 211.5 100*00 The removal of chlorine from chloranil by sulphurous acid and formation of terchlorquinone fiom the resulting products is very interesting as it will probably afford a method of obtain-ing bichlorquinone and chlorquinone in a state of absolute purity which Stadeler was unable to effect by the means he employed viz.fractional crystallisation. The substance from which the terchlorhydroquinone was obtained by sublimation Beems to be a conjugate acid but hitherto I have not been able to obtain either it or any of ita ealts in a state sufficiently pure for analysis. It is exceedingly soluble in water alcohol and ether and crystallises from the latter in large prisms. Terchlorquinone. Although as I have already stated I was unable to obtain terohlorquinone free from chloranil either by cry stallisation or * Ann. Ch. F'harm. lxix 322. STENHOUSE ON CFILORANIL. sublimation yet as I found that their hydrogen compounds admit of being readily separated it offered a method of obtain-ing large quantities of terchlorquinone in a pure state.This is the more important as Stadeler who discovered this body did not get it in sufficient quantity for analysis. The crude mass obtained in the preparation of chloranil by the process described in the first part of this paper was treated three or four times with small quantities of boiling alcohol to remove the red oil and then extracted with a considerable quantity (six times its weight) of the same solvent filtering whilst hot. On cooling the impure terchlorquinone which crystallised out was collected and the mother-liquor again boiled up with the undissolved portion left on the filter. By this means the original mixture was separated into three port ion s-I. The red oil which could be converted into chloranil by treatment with chloride of iodine solution.11. The portion almost insoluble in spirit consisting of nearly pure chloranil; and 111. The crude terchlorquinone still retaining however a considerable quantity of chloranil. In order to effect the separation of these two substances this portion (111)was added by small quantities at a time to mode-rately strong boiling hydriodic acid containing a few pieces of phosphorus. The chloranil and terchlorquinone were thus con- verted into chlorhydranil and terchlorhydroquinone which sank to the bottom as an oil and on cooling a large quantity more of the terchlorhydroquinone crystallised out. This was then col- lected washed slightly with cold water boiled with spirit and filtered from the undissolved phosphorus.The solution after evaporation to dryness at a gentle heat was exposed to the air until the phosphorus it contained had oxidised. The white crystalline mass was then powdered and submitted to sublimation in a glass vessel heated to 120"-130" in a para,& bath. The terchlorhydroquinone sublimed in brilliant plates leaving behind the chlorhydranil which is not volatiIe at that temperature. After one or two crystallisations fiom hot water (10 parts) it was found to be perfectly pure as may be seen from the results of the subjoined analysis :-I. -326 gri. mbstance gave -404grm. carbonic acid and 0050 grm. water. STENHOUSE ON CHLORANIL. I. c6 = 72 33-72 33-51 p.c. -Cl = 106.5 49-59 H = 3.4 1.41 1-70 I 0 = 32 14.89 -213-5 Sometimes the aqueous solution did not crystallise out for days but the addition of a few crystals or immersing the beaker in a freezing mixture caused it-to solidify to a mass of fine needles.It requires 160 parts of water at 15"C. to dissolve it. As terchlorquinone when digested with strong nitric acid was slowly decomposed with formation of chloropicrin and evolution of carbonic anhydride and nitrous fumes I found it advigable to effect the conversion of terchlorhydroquicone into terchlorquinone by dissolving it in hot water strongly acidu- lated with sulphuric acid and then adding a solution of acid chromate of potassium when the terchlorquinone was precipi- tated in the crystalline state. TerchZorbrompuinone.When terchlorquinone and bromine were digested together for some hours in a sealed tube at 120'-130' C. on opening the tube a considerable quantity of hydrobromic acid escaped and the residue after being washed with water was found to be very slightly soluble in hot alcohol from which it crystallised in yellow plates resembling chloranil. It contained bromine and when dried at 100' and analysed it proved to be terchlor- bromquinone C6c1,Br0,. 0350grms. substance gave ~317grm. carbonic anhydride and 0004 grms. water. This corresponds to 24-71 per cent. carbon whilst the formula C&Cl,BrO requires 24.78 per cent. Terehlorbromhydroquinone,C,~,BrO,H,. When the terchlorbromquinone was digested with hydriodic acid and phosphorus as in the preparation of chlorhydranil from chloranil terchlorbromhydroquinone was formed It was 150 STENHOUSE ON THE ACTION OF NITRIC ACID.very soluble in alcohol from which it ci-ystallised in long prisms and was readily reconverted into terchlorbromquinone by boiling with nitric acid. The analyses in this paper were made for me by rny assistant Mr. Charles E. Groves.

ISSN:0368-1769    

DOI:10.1039/JS8682100141

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

150 STENHOUSE ON THE ACTION OF NITRIC ACID. XVIL-Action of Nitric Acid on Picramic Acid. By JOENSTENHOUSE, LL.D. F.R.S. &c. FROM the conflicting statements that have been published regarding the products of the action of nitric acid on picramic acid I was led to examine into the nature of this reaction. Girard and Pugh* state that when picramic acid is treated with strong nitric acid it is coiiverted into picric acid. W oh1 er and Carey Lea,t on the contrary state that picramic acid is not converted into picric acid by the action of nitric acid. Three parts by weight of boiling nitric acid sp. gr. 1.45 were poured on one part of picramic acid which it readily dis-solved and after a few minutes a violent reaction ensued and nitrous fumes were evolved with almost explosive violence.When the action had somewhat moderated heat was applied ta the mixture and after continuing the digestion for about ten minutes the liquid was allowed to cool. This on standing some time deposited a crystalline substance in such large quantity that it became semi-solid. The crystals were drained from the acid mother-liquor on a funnel plugged with gun-cotton dried on a porous tile and after being crystallised two or three times from spirit were analysed with the following results. I. -228 grm. substance gave *286grm. carbonic acid and *028 grm. water. 11. *416grm. substance gave *524grrn. carbonic acid and *049 grm. water. # GFirard Compt. Rend. 36 421 ; Pugh Ann. Ch. Pharm. 96,82. 't WFhler Pogg.Ann. 13 188 ; Carey Lea Sill. Am. J. [2] 26 279. ON PICRAMIC ACID. 151 Theory. I. 11. Mean. = 72 34.28 34-22 34.36 34.29 0.95 1.36 1.31 1.33 %=- 56 26-67 --2 N4 -80 38.10 --05 210 100*00 From the analysis it will be seen that the crystalline com-pound above-mentioned is identical in composition with the diwodinitrophenol C6H,N,(N0,),0 of Gries s,* with which its physical properties also correspond. The mother-liquors from which the diazodinitrophenol had aeparated after concentration yielded crystals which when purified by repeated crystallisation fi-om water were found to have the composition and all the physical properties of picric acid. Subjoined is the analysis :-I. *478 grm. crystals gave -556 grm. carbonic acid and 0074 grm.water. IT. -388 grm. crystals gave 0453 grm. carbonic acid and -056 grm. water. Theory. I. 11. Mean. I -72 31-44 31-74 31.85 31.80 ‘6 H3 - 3 1.31 1.72 1-61 1.66 N3 - 42 18-34 --0 = 112 48.91 ---PI 229 1oo*oo The silver salt was also prepared by neutralising the acid with nrgentic carbonate and recrystallising. On analysis -775 grni. of the salt gave -330 grm. of argentic chloride = 32.05 per cent. of silver. Theory C,H$g(NO,),O requires 32-15per cent. In most of the experiments which I made much diazodinitro-phenol was produced and comparatively little picric acid. In one instance however this result was eiitirely reversed ; scarcely any nitrous fumep were given off and picric acid was the chief product. The effect of the action of nitric acid on picramic acid is therefore the production of diazodinitrophenol and picric acid in variable proportions probably dependent on temperature and the strength of the uitric acid employed a cir-cumstance which readily explains the discordant results of previous experimentera. * Ctriew Ann. Ch. Pharm, 113,206.

ISSN:0368-1769    

DOI:10.1039/JS8682100150

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

WANKLYN CHAPMAN AND SMITH XVIIL-Note on Frankland and Armstrong’s Memoir on the Analysis of Potable Waters. By J. ALFRED WANKLYN E. T. CHAPMAN,and MILES H. SMITH. IN the memoir lately published* by Fr a n kl an d and Ar rn strong describing the method by which they propose t.0 conduct the examination of waters as regards organic substance objections are raised to the method we have recommended for deter-mining the relative quality of water by means of the albu- minoid ammonia it yields by treatment with alkaline solution of permanganate of potash. Those objectiom are to a great extent based upoii a comparison of the results obtained in this way with the results obtained by Frankland and Arm- str on g’s method. We have therefore made a careful inquiry into the conditions to which tlie difference between the results furnished by the two methods may be referred and into the accuracy of Frankland and Armstrong’s method to be applied as a test of the results furnished by our method.In placing before the Society the conclusions at which we have arrived we must in the first place premise,? as before stated that we do not consider the complete conversion of organic nitrogen into ammonia by our method as being essen- tial to its applicability for determining the relative quality of water and that we rely simply upon the coiistancp of the ratio Between the amount of albuminoid substance in the water and the quantity of ammonia produced. The next point which we have to consider is the degree of accuracy attainable in estimating the carbon and nitrogen in the water-residue according to Frankland and Armstrong’s method.In order to enable a judgment to be formed on this point the authors gave ten examples in which known weights of known substances were dissolved in distilled water with some carbonate of soda or carbonate of lime and the residues obtained after treatment with SO and evaporation were burnt with chromate of lead. In the following tables we give * Chem. SOC.Journal (March 1868.) t Chem. SOC.Journal (December 1867.) ON THE ANALYSIS OF POTABLE WATERS. the amounts of carbon taken and the amounts obtained also the amounts of nitrogen taken and obtained and the errors applying to each experiment :-Substance taken.Carbon taken. Error. Carbon obtained. milligrm. milligrm. milligrm. milligrm. I.. 35.2 Sugar 14.82 -0.19 14-63 11.. 34.7 , 14.60 -0.74 13.86 111.. 11.4 , 4.80 -0.40 4.40 IV. 12-2 , 5-14 f-0.16 5.30 V.. 11.5 , 4-84 -0.50 4.34 TI.. 10. Urea 2.00 -0.23 1-77 VII. . 10.25 , 2-05 + 0.06 2-11 VIII.. 1.004 , 2.08 + 0.31 2.39 IX.. 20.2 , 4.04 + 0.48 4.52 X.. 25. Hippuric acid 15.08 -1.22 13.86 Nitrogen taken. Nitrogen obtained. Milligramme. Error. Milligramme. V.. .......... 2-46 + 0-08 2-54 VI.. .......... 4.66 -0.03 4-63 VII.. .......... 4.78 -1.21 3-57 VIII.. .......... 4-84 -0.16 4-68 IX.. ......... 9.42 -0.55 8.87 X.. .......... 1.95 + 0.08 2.03 From these tables it will be seen that there is a deficit of carbon in six out of the ten determinations and an excess in four of' them.The greatest error in deficit is 1.22 milligrm.; the least error in deficit is 0.19 rnilligms. ;the mean error on the six determinations being 0.49 milligrms. In ordinary organic analyses wherein 200 or 300 rnilligrms. of a substance such as sugar is taken it is possible to obtain results which are accurate to within about one-tenth per cent. equal to an absolute error of 0.2 milligrm. of carbon. When smaller quan- tities of substance are analysed in the ordinary way it is admitted that the degree of percentage accuracy attainable in-ferior. The quantities of organic substance to which the results given by Frankland and Armstrong refer are fkom lo to 35 milligrms. being about one-tenth as much as would be operated VOL.XXI. N WANKLYN CHAPMAN AND SMITH upon in an ordinary organic analysis. But the absolute emoi; as shown in the tables above is from O*2 to 1.2 milligrm. of carbon so that while operating on smaller quantities there is no corresponding reduction of absolute error and consequently Frankland and Armstrong’s method as exhibited by their own experiments does not attain to a higher degree of accu- racy than would be reached by ordinary organic analysis ap- plied to very small quantities of organic substance. The importance of this circumstance as regards the estimation of organic substance in a water will be appreciated when it is considered that the quantity of organic substance in a litre of water is seldom anything like so much as the quantities of sugar &c.operated upon in the experiments given aLs indi-cative of the degree of accuracy attainable by Frankland and Arm str o n g’s method. From the results obtained for nitrogen it will be seen that out of the six experiments the results of four are in deficit and those of two are in excess. These results apply to quantities of nitrogenous substance fully ten-fold as large as those likely to be present in a litre of ordinary water. It therefore appears to us that taking these data as representirig the extent to which this method can be depended on for the determination of the minute quantities of carbon and nitrogen in a water-residue it does not estimate quantities of nitrogen which fall short of half a milligramme.Now on turning to the table of analyses at the end of the memoir,* it will be seen that the quantity of organic nitrogen per litre (and a litre is the quan- tity of water upon which the determinations were made) is represented as ranging from 0.00 through all varieties of inter- mediate value to 0.56 milligrm. These quantities of nitrogen are however within the limits of error indicated by the experi- ments above referred to ; consequently we cannot regard these results as representing differences of quality in the different kinds of water. Having thus considered what are the capabilities of the method proposed by Frankland and Armstrong we will now proceed to &cuss the comparison which they institute between the results furnished by our method and those obtained by their own In the first place it will be seen that the diferencee * Chsrn.SOC.Journd March 1868 p. 108. ON THE ANALYSIS OF POTABLE WATERS. I55 between the results of the two methods observed by Frank- land and Armstrong range from + 0.05 to -0.52 of a milligramme amounts which as we have already shown lie within the limits of experimental error. From these differences therefore no conclusion of any kind can be drawn and we consider it to be sufficiently evident that Frankland and Armstrong’s method is incapable of testing the accuracy of the results obtained by our method as stated below. MilIigramme per litre. Albuminoid NHS. Organic Ritrogen. Wanklyn Chapman Frankland and and Smith.Armstrong. Bala Lake water .............. 0.25 0.01 Loch Katrine water ............ 0.13 0.08 Manchester water .............. 0.07 0.26 rO-06 048 Thames water as delivered in I 0.15 London by the different corn-( 0.12 panies at different dates ...... I 0.14 Lo.20 New River water ............. East London Water Company.. ,. 0-09 0.24 Caterham water.. .............. 0.00 0.07 On general grounds we are disposed to consider that the circumstance of Frankland and Arnistrong’s method being applicable only to the residue obtained by evaporation of a water is a disadvantage both as regards the time requisite for making an experiment and on account of the probability of loss of organic substance. But hi our opinion the preliminary treatment of the water with SO in order to eliminate nitrogen existing as nitrates and nitrites coniprisee a source of error of a far more serious character.There can be no question as to the complete expulsion of CO by this treatment and we there- fore pass over that part of the subject. But as the nitrogen existing as nitrates in some kinds of water is often much more than ten times as much as the nitrogen existing in organic states of combination it will be manifest that the estimation of organic nitrogen by Frankland and Armstrong’s method WANELYN CHAPMAN AND SMITH would become illusory if only a small portion of the nitrates were to escape decomposition. On referring to Frankland and Armstrong’s paper it will be seen that the process for the destruction of the nitrates and nitrites is as follows :-“ 2 litres are poured into a convenient stoppered bottle and GO C.C.of a recently prepared saturated solution of sulphurous acid are added.” . . . . “One half of this sulphurized water is now boiled for two or three minutes and unless it contained a considerable amount of carbonates 0.2 grm. of sodic sulphite is to be added during the boiling so as to secure the saturation of the sulphuric acid formed during the subsequent evaporation.” The addition of 66 a couple of drops ’’ of solution of ferrous or feriic chloride is also recommended and the water is subsequently to be evaporated to dryness upon a steam or water-bath. The authors remark further on that L“Eh an expulsion of the nitrogen of nitrates and nitrites is a remarkable reaction and could scarcely have been predicted ; indeed it takes place to a very partial extent only when a nitrate is dissolved in water and evaporated with excess of sulphurous acid in imita- tion of a natural water ; neither is the result very different when sodic chloride or calcic or magnesic carbonate is added.” We agree with the authors in looking upon a total decom- position of a nitrate by a few minutes’ boiling with a solution of sulphurous acid as remarkable.On the other hand a decom- position of free nitric acid by sulphurous acid is what we should be quite prepared to expect. When it is considered that a libre of many waters contains in solution sufficient free oxygen to generate 0.06 grm.of sul-phwic acid by oxidation of the sulphurous acid it will become obvious that notwithstanding the addition of the 0.2grm. of rrodic sulphite which is recommended in the case of those waters containing no considerable amount of carbonates there Will always be great danger of the water becoming strongly acid. The probability of the 30 C.C. of saturated solution of sulphurous acid containing sulphuric acid is also great ; there is moreover the risk of absorption of oxygen and consequent formation of sulphuric acid during the standing in the bottle and during the boiling in the flask. It is worthy of note that the addition of the “couple of drops” of solution of ferric chloride which the authors find ON THE ANALYSIS OF POTABLE WATERS.to be so efficacious in rendering the decomposition of the nitrate complete is equivalent to an addition of so much free acid. Of the six experiments given by Frankland and Ann-strong to illustrate the complete decomposition of the nitrates (pp. 96 and 97) the first one the third and fourth,are instances in which from the absence of any alkaline or earthy sulphite to take up the sulphuric acid there must necessaiily have been free nitric acid from the very beginning of the reaction. In the second of these experiments 10 C.C. of a solution of sodic sul- phite (strength unknown) were added. In the fifth experi- ment a natural water was taken but no mention is made of the amount of carbonate of lime in it.Only the sixth admits of discussion as a possible example of a complete reduction of nitrates without the charging of the water with free sulphuric acid In this experiment -01 grm. magnesia 0.1 grm. calcic carbonate 0.1 grm. sodic chloiide 0.01 grm. potassic chloride 1 drop of solution of ferric chloride 2 drops of solution of hydric sodic phosphate 0.1 grm. potassic nitrate and 15c.c. of sulphurous acid solution were taken and complete destivction of the nitrate was the result. By calculation it will be seen that the 10 milligrammes of magnesia and 100 milligrammes of carbonate of lime are equi- valent to 122.5 milligrammes of sulphuric acid. Now the oxygen dissolved in the water cannot have been less than would suffice to form about 60 milligrammes of sulphuric acid whilst 97 milligrammes of sulphuric acid would be set free by the re- duction of the 100 milligrammes of nitrate of potash.Thus we should have about 34.5 milligrammes of fkee sulphuric acid afi the final result of the reaction. It would therefore appear that in Frankland and Arm-B tr on 9’s test-experiments in which there was complete re-duction of the nitrates the circumstances were such as to give rise to free sulphuric acid as a final product. We have made experiments in which care was taken to avoid the production of this acid as an ultimate product of the re- action and have never succeeded under such circumstances in effecting a complete destruction of the nitrates. In some in-stances iron and phosphates were present in the natural waters experimented upon ;but still the destruction of the nitrates wag incomplete.158 WANKLYN CHAPMAN AND SMITH The following experiments may be cited:-Wa,ter from a pump in Great Portland Street 1litre taken 30 C.C. of a Batu- rated solution of sulphizrous acid was added and then boiled for two minutes ; evaporated on the steam-bath (one or two C.C. of a saturated solution of sulphurous acid being added four times during the evaporation). Result 14 milligrammes of HNO were left in the residue. It is to be observed that this water contains both iron and phosphates. The water from a pump in Bartholomew Lane gave a similar result. To another well-water a quantity of sulphite of protoxide of iron was added before submitting it to the action of sulphurous acid and yet the residue contained much unreduced nitra,tes.Tn another instance 100 milligrammes HNO (in the state of nitrate of potash) 300 milligrammes Ca,O.CO, 100 milligrammes KCl and about 400 milligrammes of phosphate of lime were put into half a litre of distilled water boiled with 30 C.C. satu-rated solution of sulphurous acid and evaporated to dryness. Result 55 milligrammes of HNO left undecomposed in the residue. On repeating this experiment and substituting sulphite of protoxide of iron for the phosphate of lime 69.5 milligrammes of HNO was left undecomposed. In another experiment half-litre of New River water was taken saturated with carbonic acid and then boiled cooled in an atmosphere of that gas and again boiled and cooled in car-bonic acid; in this way it was insured that the water should be frec from dissolved oxygen.15 C.C. of a saturated solution of sul- phurous acid free from sulphuric acid was then added and the whole mixture boiled for 24 minutes and evaporated to dryness. The residue was found to contain much nitric acid. (These determinations of nitric acid were made by a modification of chul z e's aluminium-process a description of which has been laid before the Chemical Society.) From all these experiments two facts are very apparent. First the operation of destroying the nitrates in water by means of sulphurous acid is a very uncertain one. Secondly the operation a6 pract'ised by Frankland and Armstrong gives rise to free sulphuiic acid in the residue We need hardly add that few organic substances will bear being heated to 100°C.with their own weight of rjulphuric acid without undergoing great decomposition. ON THE! ASALYSIS OF POTABLE WATERS. We must liere call attention to the circumstance tbat in the experiments made to determine the degree of accuracy attain- able by Frankland and Armstrong’s inethod of estimating carbon and nitrogen no nitrates were added to the water and as tliere was in all cases sufficient alkaline or calcareous sul- phite to take up any sulphuric acid produced by oxidation of sulphurous acid by t’he oxygen dissolved in the water there was no charring of the residue by sulphuric acid. In this very material condition therefore the trial experiments differ from those made to prove complete destruction of nitrates and from operations on natural water.In the case of water containing large quantities of organic nitrogen as for example actual sewage in which the amount of nitrogen would be capable of estimation by Fraiikland and Armstrong’s method we encounter another difficulty due to the presence of ammonia in the water residue. In order to arrive at the organic nitrogen it would then be necessary to make a determination of the free ammonia in the water and to deduct the nitrogen corresponding to it from the total nitrogen of the residue. Owing however to the circum- stance that ammonia would be lost by diffusion during the evaporation (even in presence of an acid) the water-residue will contain only a part of the original ammonia and consequently an error would arise in deducting the amount of nitrogen cor- responding to the original ammonia from the total nitrogen of the residue.This source of error would of course affect to some extent all determinations of organic nitrogen in waters containing free ammonia but it would become important in the case of such waters as London well water which often contains a con- siderable quantity of free ammonia. We will conclude by giving in a tabular form a number of analyses made by our ammonia-method showing the extreme constancy of the results. Each of the first five sets of analysea bracketed together waa made on the same day on the same sample of water.ON THE ANALYSIS OF POTABLE WATERS. Quantity operated Milliqrammes of Name of Water. upon. Albuminoid Ammonia per litre Litres. of the water. -1 0.070 West Middlesex water l2 0-065 { 18 0-07 ** 0-20 0.205 Southwark and Vauxhall.. 4 0*18 I 0-18 1 0.2 1 Well at Wimbledon 1 0.15 -. Q 0.16 Bishopsgate-street pump. 1 -1 0.255 Mixed with equal volume oi 4 (= 1litre2 of } distilled water . . .. mixture). 0.26 Thames water above1 0.2 1 Hampton .. 1 0.21 0205 ** -1 Manchester water all taken 2 0.06 at same date but fiom dif-1 0.07 2 ferent parts of the town -1 0-07 2 Edinburgh water taken fiom 1 0.075 a tap at the University on 1$ 0.063 18th and 19th Sept. 1867 1 0.070 London Institution.

ISSN:0368-1769    

DOI:10.1039/JS8682100152

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

161 XK-On the Action of Oxidizing Agents on Organic Cmpounda in presence of excess of Alkali. By J. ALFRED WANKLYN, Professor of Chemistry in the London Institution and E. T HEOPHRON CHAPMAN. PART I. Ammonia evolved by Alkaline Permanganate acting on Organic Nitrogenous Compounds. IT has been observed that albumin evolves ammonia when sub- mitted to the action of permanganate of potash in etrongly alkaline solutions.* Furthermore that this ammonia is per- fectly constant in quantity being strictly proportional to the amount of albumin employed and that it is not the whole but only a fraction of the total ammonia which the total nitrogen of the albumin is capable of furnishing.? On extending this enquiry to organic nitrogenous substances in general we find the action of strongly alkaline perman- ganate to be perfectly definite as will he apparent from the results to be given further on.We shall confine ourselves in this paper to a consideration of the ammonia evolved and in subsequent papers in continua- tion of the subject hope to render account of the residual nitrogen (when there is any) and of the other complementary products of the oxidation. We will first describe the mode of conducting the experi- ments. The ammonia evolved during the reaction was (as in the papers just referred to) measured by means of the Nessler test and inasmuch as this test is used with very small quantities of ammonia it was convenient to operate on very small quantities of organic matter. In order to measure these small quantities with the necessary degree of precision the following plan waa usually adopted.* Wnnklyn Chapman and Smith Journal of the Chcm. SOC.(18671,pol. v [Ser. 23 445. t Ibid. Wanklyn (1867),vol. v [Ser. 21 591. VOL. XXI. 0 162 WANKLYN AND CHAPMAN ON THE ACTION OF 100 milligrams of the substance to be oxidized were weighed out and dissolved in 100 C.C. of water thus giving a solution containing one milligram in a cubic centimetre of the solution. By measuring this solution in an accurate burette divided into tenths of a cub. cent. or by weighing out there was obtained the requisite quantity of the organic substance measured with the requisite degree of accuracy. The oxidation was usually effected as follows :-Half a litre of freshly distilled water was put into a retort then 50 C.C.of a solution of potash (,equal to 10 grams of solid potash) were added. Then about 100 C.C. were distilled over and usually found to be nmmoniacal. This having been done the further distillate was generally found to be ammonia-free. Next Borne permanganate of potash (from 0.1 to 0.5 grm.) wa8 added also a few fragments of freshly ignited tobacco pipe (Mr. Duppa's device to avoid bumping and violent boiling). Then the weighed OF measured liquid containing the sub-rstance to be oxidized was put into the retort and the distilla- tion proceeded with. The distillation was continued until the evolution of ammonia ceased or became very trifling. Blank trials having hen made with the distilled water and reagents the following substances were investigated :-I.Asparagine C,H,N,O + H,O.-The sample of nsparagine was obtained fiom Messrs Hopkin and Williams and was in beautiful crystals. It waa dissolved in water. Milligram. Substance taken = 3-65 NH obtained = 0.80 NH per cent. = 21-92. 11. Rperine Cl,H19N03.-From Messrs. Hopkin and Williams was well crystallised. It wa8 dissolved in glacial acetic acid. Milligram. Substance taken = 4.25 NH obtained = 0.23 NH per cent.. = 5.41. 1x1. ChZoride of Diamylamine (C,H ,),J%,NCI.-Prepared in OXIDIZING AGENTS ON ORGANIC COMPOUNDS ETC. 163 the laboratory of the London Institution analyaed and found to be pure. Milligrama Substance taken = 3-91 NH obtained = 0.31 NH per cent.= '7.93. The volatility of the substance rendered it difficult to get complete action which accounts fop the NH being too low. Diamylamine gives a white precipitate with the Nessler-test. IV. Amyhmine C,H,,N.-Prepared in the laboratory of the London Institution by the action of iodide of amyl on am-monia. Had a perfectly constant boiling point. Its chloro- platinate and chloride were analysed and found to be pure. It was very carefully fieed horn ammonia and gave with Nessler- test a pure white (not a brownish) precipitate. I. 11. Milligram. Milligrams. Substance taken.. .... 270 1-10 NH obtained.. ...... 0.59 0-24 NH,per cent. = 21.8. NH,per cent. = 21.8. v. D@henyLtartramide C,,H,,N,O,.-Presented to ua by &.Perkin. Milligrams. Milligrama Substance taken = 7.2 19-4 NH obtained = 0*70 1.98 NH per cent. = 9.72 10.21. The foregoing examplea are instances of total conversion of the nitrogen of the substance into ammonia a8 the following comparison shows :-Theory. Found. I. Asparagine,. ........ 22-66 21-92 11. Piperine ............ 5.96 5.41 111. Diamylamine Chloride 8.79 7.93 IV. Amy1 amine .......... 19-54 21-8 V. Diphenyl-tartramide ,. 11-33 10.21 02 164 WANKLYN AND CHAPMAN ON Tm ACTION OF The first column of figures contains the calculated quantity of ammonia which 100 parts of the substance could give if all its nitrogen passed into ammonia. The second column gives the ammonia obtained fi-om 100 parts of substance.Piperidine hippuric acid and narcotine also give up the total nitrogen in the form of ammonia on being boiled with strongly alkaline permanganate. VI. Morpl~ine,Cl,HlgNO,.-Obtained from Greville Wil-liams dried at 100" C. Milligrams. Substance taken = 20. NH obtained = 0.56 NH per cent. = 2.8. VII. Codeine C,,H,,N 0,H20.-Splendid crystab fkom Ma c-farlane of Edinburgh. Milligrams. Substance taken = 6.5 NH obtained = 0.195 NH per cent. = 3-00. VIII. Pupaverine C2,H21N0,.-From Ma c far 1a n e of Edin-burgh. Iu good crystals. Milligrams. Substance taken = 10-NH obtained = 0.22 NH per cent. = 2.2 This substance yields up the ammonia with extreme diffi- culty. IX. Strychnine C,,H,,N,O,.-From Messrs. Hopkin and Williams. In very good crystals carefully dried. Milligrams. Subatance taken = 5-5 NH obtained = 0-30 NH per cent. = 5.45. X. Iodide of Methyl-strychnine C,lH22(CH,)N2021.-From Dr. Crum Brown. Good cq-stals. OXIDIZINQ AGENTS ON ORGANIC COMPOUNDS ETC-165 Milligram. Substance taken = '7.2 NH obtained = 0.24 NH per cent. = 3.33. XI. Brucine C2,H,GN20,.-From Messrs. Hop kin and William s. Milligrams. Substance taken = 10. NH obtained = 0.46 NH per cent. = 4.6. XII. Subhate of Quinine (C,,H,,N20,),H2S0,.-From Messrs. Hopkin and Williams. It was carefully dried. Milligrams. Substance taken = 10.00 NH obtained = 0.45 NH per cent. = 4.5. XIII. Sulphate of Cinchonine (C20H2rN20)2H,S04.-From Messrs.Bullock. It was recrystallised in the laboratory of the London Institution and dried at looo C. Milligrams. Milligrams. Substance taken = 10.00 5 *OO NH obtained = 0.57 0.27 NH per cent. = 5.7. NH per cent. = 5.4. These numbers (as will be apparent fi-om the tabular atate- ment a little further on) are somewhat in excess of the theoreti- cal quantity for sulphate of cinchonine. As is abundantly evident however &om the researches which have been published on cinchonine that the formula of this sub- stance is by no means well established. The formulajust given yields 4.76 for half of the ammonia. Other formulae which have been proposed give close on 5.00 for half of the ammonia. XIV. Nicotine C,,H,,N,.-From Messrs.Hopkin and W il-liams. Milligrams. Substance taken = 1-99 NH obtained = 0.215 NH per cent. = 10.80. Nicotine haa no action on the Nesaler-teat. 166 WAMKLYN AND CHAPMAN ON THE ACTION OF XV. Naphthy lamine C,,HgN.-Prepared by ourselves from nitronaphthaline. The hydrochlorate was analysed and found to be pure. Milligram. Milfigrams. Substance taken = 12.175 4.26 NH obtained = 0.81 0.2 9 NH per cent. = Ci*dS. TXJTT? per cent. = 6-81. Naphthylamine hats no action 031 the Nestsler-test. XVI. Toluidine C,H,N.-A well-crytstallised specimen ob-tained fkom the Continent. Milligrams. Milligrams. Substance taken = 4.08 2.65 NH obtained = 0.36 0.22 NH per cent. = 8.83. NH per cent. = 8.30. Toluidine has no action on the Nessler-test.XVII. Acetate of Rosaniline C,,H,,N,C,H,O,. -Dried at 113"C. Milligrams. Milli,al.ams. Substance taken = 3.925 4-62 6 NH obtained = 0.25 0.30 NH per cent. = 6.37. NH per cent. = 6.49. In the following table under theory there are numbers cal-culated on the principle that 100 parts of the substance should give up half its nitrogen in the form of ammonia :-Theory Found NH3. NHa. VI. Morphine .................... 2-98 2-80 VII. Codeine.. .................... 2.67 3-00 VIII. Papaverine .................. 2-50 2-20 IX. Strychnine. .................. 5.09 5.45 X. Iodide of methyl-strychnine .... 3-57 3.33 XI. Brucine ...................... 4-32 4-60 XII. Sulphate of quinine.. .......... 4.56 4.50 XIII. Sulphate of cinchouine ........4.76 XIV. Nicotine.. .................... 10.49 10.80 XV. Naphthylamine. ............... 5.95 OXIDIZING AGENTS ON ORGANIC COWOUNDS ETC. 167 Theory Found NHa. NHI. XVI. Toluidine .. .......... . . .... 7.95 {i:: XVII. Acetate of Rosaniline . . ..... . 7.06 {i::; An instance of the evolution of one-third of the nitrogen in the form of ammonia is afforded by creatine C,H,N,O,. The specimen of creatine taken for the experiments was kindly given to us by Mr. Greville Williams. It was dried and analysed giving a correct result. The mean of three accordant determinations of the ammonia evolved on boiling 100parts with alkaline permanganate is 12.6. The theoretical quantity on the principle of evolution of one-third of the nitrogen as ammonia is 12.98.This result becomes intelligible when it is remembered that two-thirds of the nitrogen in creatine are present in the form of urea and that the nitrogen of urea is evolved by alkaline permanganate either as nitrogen gas or aa nitric acid. Sar-cosine the form in which the residual third of the nitrogen is contained will we believe be found to give up all its nitrogen as ammonia. In theine we have found a substance which gives up one-fourth of the nitrogen as ammonia. Found NH from 100 parts = 8.54 Theory = 8.76 (a of the nitrogen). The molecule of theine contains four atoms of nitrogen. Thebe C,H,,N,O,. Uric acid also gives up a comparatively small fraction of its nitrogen in the form of ammonia.Apparently 100 parts of uric acid yield about 7 parts of' NH,; we shall however experiment further on this substance. We subjoin determinations made on substances of which the molecular weight is unknown. 100 parts of gelatin have given 12.7 parts of NH,. *lo0 parts of casein gave 7-6 parts of NH,. 100 parts of dry albumin give about 10 parta of NH,. Q Of very doubtful purity. 168 WANKLYN AND CHAPMAN ON THE ACTION OF We mention lastly an experiment on a substance containing the whole of its nitrogen in the nitro-form viz. picric acid C,H,(NO,),HO. It gave no ammonia on distillation with alkaline permanganate but yielded nitric acid as was proved by subjequently getting abiindance of ammonia on reducing the alkaline liquid by means of ahminiurn.Before considering the general conclusions to be drawn from the experimental data contained in this paper we think it useful to give a special description of the method of research which has been followed and a discussion of some points con-nected with it. The extreme minuteness of the quantities of substance sub- jected to qua.ntitative determination is oiie of the most atriking features of this investigation and of those investigations more or less connected with it which have been published within the last year. On looking back it will be seen that the quantity of substance taken for analysis varies from one to twenty milli- grams being generally much nearer the former than the latter limit. Iu short it may be said without greatly exag- gerating that we have substituted milligrams for the grams which are ordinarily experimented on.As will have been observed we have accordingly all throughout the paper given the weighta in milligrams instead of grams. The &st point to be considered is the mode of effecting the requisite division of the substance. In one instance wherein the larger quantities were taken as for example in the instance of diphenyltartramide the substance was weighed on a bit of platinum foil and employed in the solid state. Iu every other case the substance was employed in solution. As was de-scribed at the beginning of the paper the usual way of pro-ceeding was to weigh out 100 milligrms. of substance and dissolve it in 100 cub.cent. of water or very dilute acid or alkali according to circumst'ances. The dilute solution might then if necessary be weighed out accurately to a milligram and so the substance experimented on would be capable of being divided accurately to of a milligram. If ordinary care be taken no fear need be entertained that the water used for making these dilute solutions will contain sufficient ammonia or other nitrogenous substance to vitiate the experiments. There is no great difficulty in preparing OXIDIZING AGENTS ON ORGANIC COMPOUNDS ETC. 169 distilled water of such putity as not to give so much as damilligrm. of ammonia per litre. But for making these solutions it is not requisite to use anything better than ordinarily well distilled water which made from the Londou water seldom contains more than &Q6 or 2$c lllilligram of am- monia per litre.Thus taking water of this quality we will suppose 10 cub. cent. to be employed. There would be then 10 milligrams of substance employed and an error of nlm of a milligram of ammonia introduced by reason of impurity of the water. In short a little consideration will easily show that there is no reason for apprehending the vitiation of regults in consequence of error affecting the division of the substance taken for ex-periment. By a proper system of washing the apparatus carefully cleaning out the Liebig’s condenser by distilling water through it immediately before using it for an experiment on the estimation of these minute quantities of ammonia and by carefully testing all reagents employed it is quite easy to avoid the introduction of extraneous ammonia and to obtain per- fectly regular results.It may be well to refer specially to the precautions to be taken in making a delicate testing of the purity of distilled water. The utmost freedom from all traces of ammonia is essential for instance in the case of the half-litre of water which is destined to receive the small quantity of substance the ammonia evolved by the oxidation of which is to be estimated. In citses like this it is not enough that 100 cub. cent. of the water should give no coloration with the N essler-reagent but 100 cub. cent. of the first distillate given by one litre of the water should give no reaction with the Nessler-reagent.Precise and detailed directions for the preparation and use of the Nessler-test will be found in the ‘‘Laboratory,” vol. i. p. 267. To these directions we may add that we find it convenient to have our standard solution of ammonia of such a.strength that one cub. cent. contains T&5 milligram of ammonia and that if as sometimes happens the Nessler-reagent should prove wanting in sensitiveness the addition of a little solution of bichloride of mercury will render it sensitive. Connected with the indications of the Nessler-test we have observed a point of some interest which deserve8 to have 170 WAhXLYN AND CHAPMAN ON THE ACTION OF attention directed to it. We have not met with any base except ammonia which gives the peculiar brownish coloration with the Nessler-reagent.Amylamine diamylamine and piperi- dine in very dilute solutions give white opalescence or precipitate when treated with theNessler-test ;naphthylamine toluidine and nicotine under these circumstances occasion no reaction of any kind. (Although however these volatile bases cause no colora- tion yet their presence more or less affects the tint which am- monia gives with the Nessler-test and they thus to some extent interfere with the sharpness of the estimation of ammonia.) There is therefore every reason for believing that the pro- duction of the brownish tint with the Nessler-test is quite cha- racteristic of ammonia. The degree of precision attainable in reading the indications of the Nessler-test is much greater than would be imagined at first sight.The n6cb of a ruilligrm. of NH, in 100 C.C. of liquid is a quantity very easily seen. The difference between +& and +?-so of a milligram of NH will we think be visible to most people. With practice a higher degree of precision is attainable. When instead of using 100 C.C. of water for the Nessler-test a smaller bulk is taken the indica- tions become more delicate. So small a quantity as m$5 of a milligram of ammonia may be seen in a small bulk of liquid. In short the Nessler-deteiminatios of ammonia is susceptible of the most wonderful delicacy. On referring to the results given by different substances as described in this paper it will be seen that putting nitro-com- pounds on one side organic nitrogenous substauces in general evolve ammonia on being heated to 100' C.with strongly alka- line solution of permanganates. This reaction is very general as an inspection of the very varied list of aubstances contained in this paper is sufficient to show. The compound ammonias of all kinds the amides of the acids such substances as pipeline hippuric acid creatine the natural alkaloids albumin gelatin anduric acid evolve ammonia when treated in this way. Even so tough a substance as picoline which as is well known is one of the most stubborn of orgadic compounds yields ammonia when subjected to this treatment. Except in the instance of nitro-compounds urea and ferrocyanide of potassium we have not met with any unequivocal instance of failure of an organic nitrogenous substance to evolve ammonia on being heated to 100"C.with a strongly alkaline solution of permanganate. OXIDIZING AGENTS ON ORGANIC COMPOUNDS ETC. 171 In the matter of the nitro-compounds there is be it obseived 8ome degree of resemblance between our process and the Will and Varrentrapp process. This difference however is to be noted. The Will and V a r r en tr a p p process gives irregular results when applied to nitro-compounds part of the nitrogen of such compounds forming ammonia and part not forming ammonia. Our process on the other hand does not convert any nitro-nitrogen into am-monia but into nitric acid instead being perfectly regular in its indications with nitro-compounds.It dl be understood that having converted the nitro-nitrogen into nitric acid we may subsequently reduce that acid to ammonia by means of alu- minium as exemplified in the instance of picric acid above described. On inquiring into the other peculiarities of structure which prevent alkaline permanganate evolving nitrogen of a given organic compound in the form of ammonia our attention is arrested by the example of urea which evolves none of its nitrogen as ammonia when so treated. The reason of this peculiarity is not far to seek being at once visible in the formula. Urea is in a sense a perfectly oxidized substance requiring the elements of water to transform it into carbonic acid and ammonia and if simply oxidized would exhibit a deficiency of hydrogen-there would be CO, and only 4H along with the N,.As we have seen compounds formed by the juxta-position of urea with another substance with elimination of water (as for instance creatine which as is well known is so formed from urea and sarcosine) share this property with urea and do not give up their ureic nitrogen in the form of ammonia. Possibly uiic acid furnishes so small a proportion of its nitrogen as ammonia owing to the existence of nitrogen in the weic state. And possibly the non-evolution of part of the nitrogen of albumin is due to the same cause. On turning to the early part of the paper it will be seen that whiIst amylamine diamylamine and piperidine substances all derived from homologues of marsh gas undergo total con- version into ammonia the bases toluidine napthylamine and nicotine Rubstances which are derived from hydrocarbons lower than the homologues of marsh gas undergo only a half-con- version into ammonia.This circumstance points to the CHAPMAN ON THE ESTIMATION OF conclusion that derivation from a go-called saturated hydro- carbon implies easy conversion of the nitrogen into ammonia whilst derivation from an unsaturated hydrocarbon interposes difEculties in the way of conversion into ammonia. The conversion into ammonia of only half the nitrogen of ao many of the natural alkaloids is an interesting fact. Some light is thrown upon it by the example of narcotine which although it gives up all its nitrogen as ammonia gives it up only slowly and by dint of putting back the distillate; it doubtless also yields part of its nitrogen as methylamine in the first instance.A careful examination of strychnine has disclosed a somewhat similar state of matters in the case of that alkaloid. Apparently the missing half of the nitrogen in strychnine passes pro-visionally into the state of some volatile alkaloid. As we said at the commencement we reserve the treatment of the residual nitrogen for a future occasion.

ISSN:0368-1769    

DOI:10.1039/JS8682100161

出版商:RSC    

年代:1868

数据来源: RSC

摘要:

172 CHAPMAN ON THE ESTIMATION OF XX.-Note on t7he Estimation of Nitric Acid in Potable Waters. By ERNESTTHEOPHRON CHAPMAN. A METHOD for the estimation of nitrogen existing as nitrates and nitrites in potable waters should be at once delicate and simple. I have been using for some time past a method which in common with Fr. Schulze’s method depends on the reduc- tion of the nitric acid to ammonia by means of aluminium in alkaline solution. Instead however of using a quantity of aluminium known to liberate a known volume of hydrogen when treated with an alkali and observing the diminution caused by the nitric acid I measure the ammonia formed either by the Nessler-test or by titration. The first thing to be done is to obtain some idea of the quantity of nitrate contained in the water.For this purpose boil a little of it in a flask with caustic soda free from nitrate until a sample of it does not colour the Nessler-test cool and then introduce a strip of aluminium foil. As soon as this is dissolved decant the liquid into a test-tube add Nessler-test NITRIC ACID IN POTABLE WATERS. 173 and note the colour produced ; should it be very dark or should a precipitate be formed either very little of the water (10 C.C. -25 c.c.) must be employed or a tolerably large quantity and the ammonia produced determined by titration. If on the other hand the colour is very pale 200 C.C. will be a convenient quantity t,o employ ; for common waters such as those supplied to London 100 C.C. is ample.For waters of the Loch Katrine class 300 C.C. would be a convenient quantity. Having learned what sort of water is in hand measure off the volume indicated by the preliminary experiments-introduce it into a non-tubulated retort and add 50 to 70 C.C.of a solution of caustic soda containing 100 grms. of soda in the litre. This soda must be free &om nitrates. If very little water has been taken add some distilled water. The content8 of the retort are now to be distilled until they do not much exceed 100 c.c. and until no more ammonia can be detectedin the distillate by the Nessler-test. The retort is now cooled and a piece of aluminium introduced into it (foil will answer very well with dilute solutions but I much prefer thin sheet aluminium in all cases).The neck of the retort is now inclined %L little upwards and its mouth closed by a cork through which passes the narrow end of a small tube filled with broken-up tobacco pipe wet either with water or better with very dilute hydrochloric acid fiee from ammonia. This tube need not be more than an inch and a half long nor larger than a goose-quill. It is con- nected with a second tube containing pumice-stone moistened with strong sulphuric acid. This last tube serves to prevent any ammonia from the air entering the apparatus which is allowed to stand in this way for a few hours or over night. The contents of the pipe-clay tube are now washed into the retort with a little distilled water and the retort as adapted to a condenser the other end of which dips beneath the surface of a little distilled water free from ammonia (about 70 to 80c.c.).* The contents of the retort are now distilled to about half their original volume ; the distillate is made up to 150 c.c.; 50 C.C.of this are taken out and the Nessler-test added to them. If the colour so produced is not too strong the 9 Condensers are very apt to contain a trace of ammonia if they have been standing all night and'must therefore be washed out with the utmost care. I prefer to distil a little water through them until I can no longer detect ammonia in the distillate. CHAPMAN AND SMITH ON THE ACTION OF estimation may be made at once; if it is the remainder of the distillate must be diluted with the requisite quantity of water.Should it be desired to determine the ammonia by titration a much larger quantity of the water must be employed. Ealf a litre or a litre must be evaporated down to a small bulk and this small bulk treated in exactly the same manner except that the distillate is received in dilute standard acid instead of water. With regard to the precision attainable by this method all that I can say is it is limited by the precision of the determina- tion of ammonia by Hado w’s quantitative method founded on the Nessler-test. If therefore the quantity of nitric acid be above -0005 grm. we may be sure of our result to 5 per cent. if the quantity be under this the estimation will be somewhat less sharp. Even *00004grm. nitric acid may be detected and an estimation made by this method. I need hardly remark that the purity of the reagents em-ployed must be ascertained by making blank experiments. have found common caustic soda sold in lump to be free from nitrates. Should the caustic soda contain nitrate it may be purified by dissolving a small quantity of aluminium in its cold aqueous solution and then boiling to expel ammonia.

ISSN:0368-1769    

DOI:10.1039/JS8682100172

出版商:RSC    

年代:1868

数据来源: RSC