128 2311.-On the Reaction between Xodium Thiosulphate and Iodine. Estimation of Manganese Oxides and Potassium Bichromate. By SPENCER UMFREVILLE PICKERING B.A. Brackenbury Scholar of Balliol College Oxford. 1~ estimating the amount of manganese dioxide contained in various mixtures of the oxides of manganese as described i n a former paper, (this Journal 1879) a modification of Bunsen’s volumetric method was employed which although it had not been previously noticed by other chemists appeared sufficiently obvious to require no special mention. Instead of boiling the oxides with hydrochloric acid and collecting the evolved chlorine in a solution of potassium iodide (F. Mohr, Lehrb. d. Chern. Analyt. Titrirmethode 4 aufl. 278) or digesting them at 100’ in a stoppered bottle containing hydrochloric acid and the iodide (ibid.%l) the sample to be analysed was transferred to a beaker containing a large excess of this latter solution a small quantity of acid added and the liberated iodine determined by directly running into t)his mixture a standard solution of sodium thiosulphate. The oxides if in a state of fine powder and especially if recently pre-cipitated were found to be dissolved readily by very dilute acid in the presence of the iodide. On comparing analyses made in this manner with those made ecccording to Bunsen’s directions it was found that the former nearly always yielded slightly higher results than the latter this fact led to the following investigation of the various circumstances which influ-ence the reactions involved.PART I.-Reactions betwee@ Xodium Thiosulphate and Iodine. $ 1. Amount of Xulphate formed.-It is well known that in the re-action between iodine and sodium thiosulphate if the liquid be warm, then besides the action-I. I + ZNa,S,O = Na2S406 + 2Na1, another action also takes place resulting in the formation of a sulphate according to the equation-11. 41 + Na2S,03 + 5H,O = 2NaHS04 + 8HI. Notwithstanding the statements of Rose (abid. 270) and others a qualitative test sufficed to show that some sulphate is formed even a PICKERING ON THE REACTION ETC. 129 ordinary temperatures and although the amonnt was not large it was sufflcient to admit of its being determined quantitatively.* The sodi urn thiosulphate which contained *013704 gram N&S2O3 per c.c.contained also some sulphate and this had to be deter-mined by blank experiments. The following numbers were ob-tained :-100 C.C. of the sodium thiosulphate yielded . . . . . . . . . . . . . . . . . . . A *000336 added to an extremely slight excess of iodine at 20" C yielded . . . . . . 100 C.C. of the sodium thiosulphate hence an amount of sulpliate corresponding to *000108 gram BaS04 is formed in the oxidation of *013f04 gram Na,S203 by iodine a t 20" C., or the number of molecules of iodine reacting according to the equz-tions I and I1 is 1 and 46.6 respectively.? 9 11. Efect of Temperature.-In order to obtain Some ides as to the effect of t,emperature on these two reactions a series of' experiments was performed in each of which 10 C.C.of an iodine solution (contain-ing ,020672 gram of iodine per c.c.) kept in a bath a t a definite tern-perature was titrated by the thiosulphate. The results thus obtained are given in Table I. Blank experiments were first performed to awertain whether any appreciable loss of iodine took place from volatilisation during the titration of the warm solution. The iodine solution was heated to about 50" 0. (the highest temperature employed in any of the experiments) in a stoppered bottle; some water was then r u n into it from tbe burette the stopper replaced and the solution cooled. It was then titrated ; no loss of iodine has taken place during these operations. Before the addition of the last drop of the thiosulphate solution the liquid was in every case allowed to attain the temperature of the room.* Owing to this formation of sulphate the neutral solution of thiosulphate becomes distinctly acid after the addition of the iodine as would be inferred from equation IT. I n order that the thiosulphate used in these experiments should be as pure a~ pos-sible it was recrystallised three or four times and before the addition of the iodine to its solution both these reagents were mixed separately with a few drops of barium chloride solution and allowed to stand for 24 hours so t,hat any trace of sulphate present in them might be detected and eliminated. That the sulphate formed at ordinary temperatures in the reaction under discus-sion could not have been entirely or even principally due to any sulphite present in the thiosulphate is shown 5y the fact that its amount is considerably reduced by a reduction of tempemture.t Molecule for molecule potassium thiosulphate was found on oxidation Kith iodine to Jield nearly the same amount of sulphate as the sodium salt 130 PICKERING uN THE REACTION TABLE I.-Showirg the In$uence of Temperature on the Actions. Iodine solu-tion t,aken. 10 C.C. ) . . . . . . . 7 ) . . . . . . ,) ) ,) . . . . . . . ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temperature. Thiosulph ate required. 2 *352 c.c.* ii :$} 18 *17 C.C. 18 *a07 c.c.Q 18.32 C.C. 18.35 c.c.* : :} 18 -42 C.C. 18 -426 c.c.* 18 '47 C.C. (see Table VJ) 18.485 c.c.* 18 '497 c.c.* 18.51 .18 *52 18 *51 C.C. 113 -816 c.c.* 18.50 1 Weight of iodine reacting to form tetra-thionate. - ~~ 0 96 *lo 96 -33 97 -00 97 -17 97 *63 97 *65 97 *go 98 '03 98 *06 98 -16 100 Weight of iodine react-ing to form s ulpha te . 100 3 -90 3.68 3 .oo 2 -83 2 -37 2 -35 2'10 1 '97 1 *94 1 -84 0 The above experiments are essentially a reproduction of those made by Wright in 1870t. (Cltem. News 21 103). He however does not appear to have recognised the formation of any sulphate at tempera-tures below 28" C. (the iodine not being in excess) and its formation above this temperature he attributes to the oxidising action of the iodine on the tetrathionate formed in the first stages of the reaction, and not to the direct oxidation of' the thiosulphate according to equa-tion 11 page 128.The following considerations however appear to render such dr view untenable :-(1,) I€ the sulphate were formed by the oxidation of the tetrathio-nate then when the iodine is added to the thiosulphate we should expect that less sulphate would be formed than when the thiosulphate is added t o the iodine for in the latter case some tetrathionate is in the presence of excess of iodine throughout the reaction whereas in the former case no excess of iodine is present at any time. The experiments given in Table VI however show that the reaction is the same whether the iodine be added to the thiosulphate or the thio-sulphate to the iodine. * Interpolated. t. Wright did not make any direct determination of the sulphate formed his ex-periments consisted chiefly in ascertaining the relative amounts of iodine required for the oxidation of a given quantity of sodium thiosulphate between the temperatures of 16" and 92" C BETWEEN SODIUM THIOSULPHATE AND IODINE.131 (2.) In the case of the sulphate being formed by the oxidation of the tetrathionate the results of the experiment at 10" C. (for instance) given in Table I would be represented by the following equations :-''P6 [2Na2S,03 4- I = Na$3,Os + 2NaIl. 1.94 16 [NaZS406 + 71 + 10H,O = 2NaHS04 + 2H2S04 + 14RI], from which taking the titres of the various solutions and the quan-tities used as given above we find that during the ten minutes allowed for the reaction -00352 gram of iodine has been used in oxidising some of the sodium tetrathionate formed into sulphate the total quantity of tetrathionate formed being 0.216 gram.In order to ascertain whether the oxidation of sodium tetrathionate does in reality take place at this rate several portions of a pure sample of this salt weighing -216 gram each were dissolved in water and 10 C.C. of the iodine solution added to each.* The residual iodine was subsequently determined after various intervals of time had elapsed, the temperature being kept constant at 10" C. The following were the results obtained :-2NazSz0 + I = Na$?&O + %"I. C.C. gram. After 10 mins. 0.91 of iodine or *000207 I per 10 mins. had disappeared. , 19s hrs. 0.38 , , -000066 , 9 7 9 , 24k , 0.40 , , -000057 , 9 7 9 , , 4 6 i , 0.70 , , -000052 , 9 ) 3 , , 69 , 1-05 , , -0000525 , 9 9 7 The amount of oxidation is thus seen to diminish considerably as the action proceeds,+ but taking the first experiment which gives by far the highest rate (possibly due in part to the difficulty of measuring such small differences as *01 c.c.) we find that this rate is less than one-twelfth of what it should be if the sulphate were formed (in the reaction of iodine on thiosulphate) by the oxidation of the tetrathio-nate and not by the direct oxidation of the thiosulphate.§ 111. Bfeect ofDiZutiorz.-The experiments given in Table I1 were performed in order to ascertain whether the amount of water present had any influence on the relative proportions in which the two above-* This is of course a great exaggeration of the conditions existing in the experi-ment in Table I for here we have both the maximum amount of tetrathionate and also the maximum amount of iodine present a t the same time whereas in the ordinary determination when either of these substances is at a maximum the other is a t a minimum.t This diminution in rate is not due to the presence of the sodium sulphate formed for on adding excess of this substance the rate of oxidation was not found to be appreciably altered 132 PICKERING ON THE REACTION Iodi;yeyon mentioned reactions (page 128) take place. Dilution is here seen to cause a small increase in the amount of sulphate formed less of the thiosulphate being required according as the quantity of water present is greater. Its effect however is very slight for in order to make the experiments strictly comparable a small correction must be applied since it was found that a quantity of iodine (*000@42 gram) equivalent to -004 c.c of the thiosulphate was requisite for every 50 C.C.of solution to give a visible coloration with starch. TABLE 11.-Showing the Efect of Dilution,. Excess of added. Thiosulphate solution required. Iodine solution taken. Water added. None 50 C.C. 97 9 9 100 C.C. 7 1 2 150 C.C. ,9 J 3 1, 9 7 91 200 C.C. 250 C.C. 300 C.C. ? 9 9 , 9 9 Thiosulphate solution required. ~ ~~~~ 18 *47 C.C. (see Table TI) 18.47 18.467 C.C. 18 -47 18 '46 18-45 18.457 ,, 18 '46 18 -46 18.42 I I 1 18 '467 18 -44 18 -45 18 *P3 18.42 18 *44 18 -43 18 -42 18 -43 18 '41 18 -42 18'435 ,) 18.435 ,, 18.435 ,) 18.42 ,, Thiosulphate corrected for final react,ion.-18.47 C.C. 18.471 ,) 18.465 ,, 18'44'7 ,) 18-451 ,) 18.445 ,) 18.444 ,, 9 IV. Amount of Potassium Iodide present.-The experiments given in Table I11 show that the presence of an excess of potassium iodide over that necessav for the solution of the iodine has no effect on the quantity of thiosulphate used. TABLE TII.-Showing the efect of Excess of Potassium Iodide. I I l- I-- --l o C.C . , . ,y None (see Table VI) 1 gram 2 grams 3 Y, Y 1 YY 1, 18 *47 C.C. 18.46 ,, 18.47 ,, 18'47 ,, l 8 . 4 i ,, 18-46 ,, 18-47 y BETWEZN SODIUM THIOSULPHATE AND IODINE.133 5 V. Iizfluence of Time.-A series of experiments were next per-formed with a view of ascertaining whether in n dilute solution of iodine in potassium iodide the amount of the former diminished appreciably on being kept for various lengths of time. The solutions to be titrated were placed in stoppered bottles inverted in a vessel containing starch water so that any leakage might be detected. Some of them were kept in the dark others in diffused daylight. In both cases however no sensible diminution in the amount of free iodine present was found to have taken place within four days,* that time being the extreme limit allowed in any of the experiments. The re-sults are given in Table IV. Iodine solution added. 10 C.C. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TABLE IV.-SShowing the InfZuence of Time. Water added. Interval between dilution and titration. None $ hr. 1 hr. 2 hrs. J 9 Y, Thiosulphate required. 18.45 to 18.42 C.C. (see 18 *44 C.C. 18-45 ,, 18-45 ), 18 *43 ,, 18.43 ), 18.44 ,, 18 -42 ,; 18-42 ,, 18 *45 ), 18-45 ,) 18-42 ,, 18.44 ,, 18.43 ), 18.43 ,, 18-44 ,, 18.42 ,, Table 11) 9 VI. Ii$uence of Hydq-ochloric Acid.-In order to ascertain whether the reactions under consideration were influenced by the presence of hydrochloric acid the amount of iodine liberated by treating potassium iodide with the latter substance had previously to be determined.* It is scarcely necessary to mention that precautions were taken for detecting any change in the titre of the standard solutions ; the iodine and thiosulphate were com-pared together and the latter compared with a standard solution of potassium bichromate at least once in every 24 hours. I n order to facilitate comparison of results in the tables given the actual amount of thiosulphate employed is reduced to what it would hare been had the solutions remained unaltered throughout. The addition of a little sodium hydrate greatly increases the stability of the thio-sulphate (Harcourt and Esson Phil. Truw. 5 156 p. 205) 134 PICRERING ON THE REACTION 5 C.C. of the acid employed (density 1.156 at 20" C.= 31.6 per cent. HC1) when added to 0.5 gram potassium iodide dissolved in 45 C.C. of water were f'ound-To liberate at once To have liberated after . . 2 4 hours *00515 ,, (mean of many expts.) -00123 gram I > , another 24 hours *OW515 ,, 9 7 9 9 9 -00504 ,, 7 9 9 -00526 ,, ,Y 9 9 7 -00493 ,, 7 7 9 7 -00515 $, 9 7 7 *00515 ,, Mean -00512 gram I per 24 hours. To have liberated after . .96 hrs. -01725 gram I = -00431 gram I per 24 hours. With the quantities of potassium iodide and water above mentioned the iodide liberated at once was found to be proportional to the hydro-chloric acid added when the quantity of this latter was varied between 1 and 15 c.c. though with larger quantities than 5 C.C. the final action becomes rather uncertain.The results obtained by titrating definite quantities of iodine solu-tion in presence of hydrochloric acid are given in Table V and show that the acid has no influence on the relative proportions in which the reactions I and 11 page 128 take place. In these and other similar experiments the thiosulphate was' run into the iodine solution as soon as it had been mixed with the acid and cooled down to 20" C., 10 minutes being allowed in every case for the reaction to complete its elf. TABLE V.-Slzowin,g the InfEuence of Hydrochloric A cizl present. Iodine soh-tion taken. 10 C.C. , ) 7 . . . . . . . . ,) ~~ ~~ HCl added. ~~ Thiosulphate required. 18 '47 C.C. (see Table VZ) 18 -50 C . C . 18 53 ,) 18.53 ), 18'56 ,) 18'57 ), 18'58 ,, 18'58 ), 18-69 ,) 18'68 ), Correction for I liberated by HCI.None .02 C.C. thiosulphate *065 C.C. ,) '065 ,, Thiosulphate required cor-rected. 18-47 C.C. 18'48 ,) 18.465 ,, 18-465 ) 7 18.4'7 ,) 18.46 ,, 18-47 7, 18-47 ,) 18-47 ,, 18.4 BETWEEN SODIUM THIOSULPHATE AND IODINE. 135 § VII. Iodine added to excess of TlLiosulphnte.-The results given in Table VI show that no difference in the amount of sulphate formed is made by adding the iodine solution to excess of thiosnlphate and sub-sequently determining t)hat excess by means of a standard solution of iodine instead of adding the thiosulphate to the iodine." TABLE TI. Thiosulphate added to Iodine. Iodine solution taken.~~ 10 C . C . . . . . . . . . . . 7 . . . . . . . . ) . . . . . . . . . . . . . . . . . . . . . . Thiosulphate required. 18 *47 C.C. 18.47 ,, 18.47 ,, 18-47 ,, 18-46 ,, lS.47 ,, Iodine added to excess of Thiosulphate. Thiosulphate required. Iodine solution taken. 10 C.C. ,, 2, 7, 7 7 7 7 9 2 I 18.46 C.C. 18'47 ,, 18.47 ,, 18-46 7) 18.47 ,, 18.47 ,) lS.47 ), The case however is different if hydrochloric acid is present. The results obtained under these circumstances are given in Table VII, TABLE VI1.-Iodine and Acid added to Thiosulphate. HCl Thiosulphate added. required. 18'47 C.C. (see 1 Table VI) 5 , 18'43 Correction for I liberated by HC1. None *02 C . C . thio- { sulphate } '045 C.C. thio-sulphate *065 C.C.thio- { sulphate } '09 C.C. thio- { sulphate } '11 C.C. thio- { sulphate } Corrected thiosul-phate. -18.47 C.C. 18.45 ,, 18 -435 ,, 18-41 ), 18'38 ,, 18 *325 ,, Iodine in-dicated by the tliio-eulphate used. 100 *@O 99 -89 99 -81 99 -67 99 *51 99 '22 * Finkener probably ascertained this fact see Rose Handb. d. Anal. Chem. 6 aufl., roil Finkener 2 937 136 PICKERING ON THE REACTION and show tbat an increase in the amount of sulphate formed is occa-sioned by the presence of the acid the greater being that increase according as the quantity of acid present is greater. In these experi-ments since the acid liquid could not be added to excess of the thio-sulphate without causing its decomposition the iodine solution mixed with the acid was added to nearly the necessary volume of thiosul-phate (18.3 c.c.) and the additional quantity required run into this mixture.It is also to be noticed that the individual experiments in this series are less concordant among themselves than in any other series. PART 11.-Estimation of Manganese Oxides and Potassium Bichromate. 9 I. Valuation of ManJganese Oxides.-In valuing the oxides of manganese according to Bunsen’s method three different forms of apparatus were employed :-(1) The oxide was boiled in a small flask fitted with it thistle funnel the chlorine evolved being absorbed in three other flasks containing potassium iodide solutions ; the flasks were fitted with india-rubber stoppers and a current of air was drawn through the whole apparatus during the experiment.(2) The oxide was boiled with the acid in a small retort the neck of which was bent down and fitted by means of an india-rubber plug into the first of three U-tubes containing potassium iodide solution ; no current of air was employed. With this and also with the first apparatus no chlorine ever passed beyond the second absorption vessel. (3) The ebullition was performed in a small retort as in the second case the neck of which passed into a larger inverted retort containing potassium iodide solution. With the first two arrangements identical results were obtained ; xith the third one however a small loss of chlorine was found to take place owing to there being but one vessel for its absorption. In most cases apparatus (1) was employed as being found the most con-venient of the three.Traces of chlorine were found to be retained by the acid liquid even after prolonged boiling and these were estimated by pouring some solution of potassium iodide into the flask before disconnecting it from the absorption vessels and determining the iodine liberated in this flask separately from that liberated in the others.* A given number of minutes was allowed for the ebullition cooling and deter-mination of the iodine the amount of iodine liberated by boiling a given quantity of hydrochloric acid alone being determined in each special case by blank experiments. * It was ascertained that neither the acid nor the oxides employed contained any traces of iron BETWEEN SODIUM THIOSULPHATE AND IODINE.137 The results thus obtained with a sample of pure artificial manganese oxide are given in Table I together with the numbers obtained with the modification of Bunsen’s method described on page 128 (due cor-rection being also made in this case for the iodine liberated by the acid 5 C.C. of which were employed in each determination). Provided the acid used in Bunsen’s method be not diluted both methods yield practically identical result’s ; a small quantity of water however, caused an appreciable diminution in the amount of chlorine liberated, and this diminution becomes greater as the amount of water added is increased but in a decreasing ratio. From 0.2 to 0.3 gram of the oxide was used in each analysis this quantity requiring 20-30 C.C.of the thiosulphate. TABLE I.-Analyses of Manganese Oxide. Method employed. ~~ ~~ ~~~~~ ~ ~ Modification of Bunsen’s method . . . . . Y ’ > Y 9’ > 9 ,Y ?’ 3 27 Bunsen’s method; using 10 C.C. HC1 and no water 1’ 7) 99 Y 7 ) 9 , 9’ 7 9 ,7 Y > Y Y ’ 9 Ditto ; using 10 C.C. HCl and 5 C.C. Ditto; using 10 C.C. HCI and 10 C.C. water water Y’ 2 7 9 Ditto; using 10 C.C. HCI and 15 C.C. Ditto; using 10 C.C. HC1 and 20 C.C. water water ’2 > 9 7, ” Y9 >7 3’ ,J ’, Ditto; using 10 C.C. HC1 and 30 C.C. water 9’ 9 Y’ Y ’> ’Y Ditto ; using 10 C.C. HC1 and 50 C.C. water 9 9 ,Y 9 , Y Y Y Y ’7 Y >> Percentage of avail-able oxygen found. ~ 6 *938] 6922 1 6 -640 } R *932 6.923 { 6 -936 J 6 *922) 6 *938 I 6 a934 k6 6.918 I 6 *938 J 6 *900* 6 2379 6 *881 6.864* 6 -872 6 -840 6.850) 6 *824 6.831 6 -801 [ 6 $14 6.832 I 6 %29 J -Relative quantity of available oxygen found.100 -00 99 *97 99 -54 99 -24 99 *01 98 -904 98 -56 98 -30 * Interpolated by means of a curve 138 PICRERING ON THE REACTION 5 11. Analyses of Potassium Bichromate.-A series of experiments, similar to those just described were performed substituting potas-sium bichromate for the manganese oxide. The results are given in Table IT and are found to agree with those in Table I ; they show, however that in the case of potassium bichromate the d-iscrepancy between results obtained by the two methods is greater than with manganese oxide and thus even when the acid employed in Bunsen's method is undiluted a notable deficiency of chlorine is observed.TABLE 11.-Analyses of Potassium Bichromate. Method employed. Modification of Bunsen's method Y I f 9 > > 3 )> ) > >) > )> >) 7 ) - - - ~ -Bunsen's method ; using 10 C.C. HCl and no water 97 9 ) Y ) 1 ) >) >) ) 9 > Y 9 ) > f Ditto ; using 10 C.C. HCI and 3 C.C. water > 7 >> 9 ) 9 ) 9 93 Ditto; using 10 C.C. HCI and Ditto; using 10 C.C. HC1 and Ditto ; using 10 C.C. HC1 and Ditto ; using 10 C.C. HCl and Ditto ; usiiig 10 C.C. HC1 and 5 C.C. water 7 C.C. water 7) 7, 10 C.C. water 15 C.C. water 15' C.C. water 77 7 ) 7) ) > Thiosulphate required. -35 *04 :g :: 135 *046 C.C.35.05 1 35 .oa' 35 -04 J 34 -847 34 *83 1 34 *835 C.C. 34.83 34 -83 34.84 1 ; I;:{ ;t :y 1 34 9'6 } 34 *755 C.C. 34.69" C.C. 34 -66 34 -66 34 '625" C.C. 34.595* C.C. 34 -59 34 -66 C.C. I 34 -58 Relative quan-tity of thiosul-phate used. 100 40 -99 -40 99 *17 98 *9s4 98.90 98 .SO 98 '713 98 90 0 111. Loss of Chloriice.-In order to ascertain whether the well-known formation of hydrochloric acid in chlorine water at ordinary * Interpolated by means of a curve BETWEEN SODIUM THIOSULPHATE AND IODINE. 139 temperatures was sufficiently increased at 100" C. to account for the low results obtained by Bnnsen's method some chlorine water was heated for various times at different temperatures either in stoppered bottles or in sealed tubes and the numbers thus obtained are given in Table 111 the chlorine water here used not being saturated but con-taining 0.002192 grain C1 per C.C.All the experiments except the last one were performed in weak diffused daylight ; the amount of hydrochloric acid formed however does not appear t o be increased to any great extent by the action of light. TABLE 111.-Showing the Loss of Chlorine produced by Heating Chlorine Water. Temperature. Before heating 2, 7, Heated :o 100°C and cooled At 180" C. slovvl y 7, 3 , 7 >, ,, 7 -_-At 62" C. 7 41 7, Y 20 ,, in the dark ,> Time. ----bbout 4 hr 9 for Q hr. for 1 hr. 9 , ,> Y9 9 , for 2 hrs. for 1 hr. €or 24. hrs. -9 , Y, Thiosulphate required.24 -99 C.C. 25 -98 26.02 24.94 25 *04 g2 :;:} 24 *435 C.C. 23 *61 23 -58 23 -68 22 *37 C.C. 24.81 C.C. 25.28 ,, 24.62 ,, 24.89 ,, Loss of free chlorine espe-rienced. -3 *90 p. c. 6-06 ,, 9.16 ,, 14.00 ,, 4.63 ,, 2.80 ,, 0.22 , per hr. 0 -18 9 9 , These experiments show that a very considerable loss of free chlo-rine takes place at loo" and idso at lower temperatures," being quite sufficient to account for the loss experienced in Bunsen's method, especially a the chlorine is then in a nascent condition; and that * Some chlorine water was heated over mercury that the oxygen liberated might, be collected ; in place however of this gas being evolved the mercury was oxidised, an oxychloride being probably formed.When some mercury wa8 shaken with strong chlorine water it was converted a t once into a fine grey powder which on subsiding, left the liquid quite colourless. Lead immersed in chlorine water became oxidised immediately the metal and especially the parts of the glass vessel near it becoming coated with a film of what appeared to be lead dioxide 140 BLYTH APPARATUS FOR THE TREATMENT OF SUBSTANCES. this is the true explanation of this loss is supported by the fact that the results obtained are lower in proportion as more prolonged boiling is necessary to liberate the chlorine i.e. the results are lower as the acid used is more dilute and also lower in the case of potassium bichromate than with manganese oxide. Nor can any of this loss be considered as mechanical for if it were so it would be extremely improbable that an entire change in the apparatus used should not alter its amount and impossible that no loss should take place when undiluted acid is employed for the solution of the oxide. § IV. ConcZusioib.-It hence appears that Bunsen’s method is capable of yielding accurate results only in the case of manganese oxides, using the strongest acid. The modification of it here invest’igated has the advantage of being not only accurate but far more expedi-tious and less troublesome than the original method ; a smaller quan-tity of acid being required (2 to 5 C.C. for a determination) and the cor-rection due to the iodine liberated by this acid being determinable with greater ease and certainty than in Bunsen’s process. Unfortunately, however it has the disadvantage of not being applicable to manga-nese ores since the ferric oxide present in them would also liberate iodine from the potassium iodide and therefore in these cases Bun-sen’s method must be employed using all the precautions here indicated