摘要:

THE ANALYST. SEPTEMBER, 1896. PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS, COMMERCIAL PRUSSIAN BLUE. BY ERNEST J. PARRY, B.Sc., F.I.C., AND JOHN HENRY COSTE, F.I.C. (Read at the ileetiiag, June 3, 1896.) IN G. Hurst’s well-known book, “ Colours, Oils, and Varnishes,” a scheme for the analysis of colours containing Prussian blue is given as follows : ‘‘ A known weight of the sample should be boiled in caustic soda until the blue is completely decomposed. The mixture is then filtered and the brown residue well washed with hot water until it is quite free from alkali. Next, it is treated with hydrochloric acid until the brown oxide of iron has been dissolved. Lastly, it is filtered, and to the filtrate is added ammonia in slight excess. The precipitate of oxide of iron thus formed is filtered off, washed with water, dried and burnt in a crucible, and then weighed.The weight of the oxitlc of iron wmltiplied by 2.312 givcs the umo1mt of Prmsinn b1u.c in the sample takcic.’’ As a matter of fact, under these conditions Prussian blue yields about 28 per cent. of Fe20y, so that estimations based on Hurst’s statement would fall short of the truth by at least 35 per cent. Hurst is here evidently under the impression that all the iron in the pigment is precipitated by treatment with alkali, whereas, of course, all in the form of iron-cyanogen radicle is lost in the filtrate as alkaline ferrocyanide. The following is given as the complete analysis* of a sample of Chinese blue, one of the best varieties of Prussian blue: ... Water ... ...... 4.487 Oxide of iron ... ... ... 52.055 Cyanogen ... ... ... ... 43.508 ~ 0 0 ~ 0 0 0 The method here employed is evident. Briefly, we presume it involves the determination of the moisture and the ash. Since the pigment contains no oxygen, even if the ash were entirely oxide of iron, the iron present is only 70 par cent, of the amount of oxide, and the cyanogen, evidently taken by difference, would be corre- spondingly higher. We have found, and Dr. Williamson? in every case found, that alkali was present in so-called Prussian blue, not merely as alkaline salts not washed away completely, but as an integral part of the cyanide. This alkali does not figure in the above analysis. * These figiires add up to 100.05, evidently by a clerical error. I- C’hem. Xoc. Mem., 1845-46, 125.226 THE ANALYST.As we are frequently called upon to examine commercial samples of Prussian blue, and of pigments cont.aiaing this substance, we have considered it desirable to make a more complete examination than the method given above would afford of specimens of blue of known origin, with a view to getting some more definite data for basing conclusions as to the purity of samples of unknown make. I t appeared to us that, more especially in the case of complex pigments ground in oil, the determination of the nitrogen would give valuable information as to the amount of Prussian blue present. Dyer has shown (Cl~em. SOC. Jozir., 1896, 811) that ferrocyanides and ferricyanides yield the whole of their nitrogen as ammonia when treated by the Kjeldahl-Gunning- Arnold method.We found that potassium ferrocyanide boiled with sulphuric acid for a few minutes gave up all its nitrogen as ammonia. A determination of the water of hydration (lost at 180" to 200" C.) was also made. The results were : Found. Theory. N ... ... ... 19-64 19-95 H,O ... ... 12-79 12-71 We therefore felt justified in assuming that equally correct results would be obtained in the case of tbe allied compounds cont,ained in commercial Prussian blue, an expectation which was fully justified by the agreement of the nitrogen determina- tions with the other figures obtained. The following is briefly the scheme of analysis adopted by us : The moisture is determined at 100". The water of coinbination-if the term is here justifiable-can scarcely be determined unless the blue is '' burned " and the water collected, as we made several attempts to drive off the water in a current of hot air, gradually raising the temperature until at 230" C., when water appeared to be still coming off, the pigment was decomposed with the formation of ferric oxide.Dr. Williamson, in his classical paper, " On the Blue Compounds of Cyanogen and Iron,"" determined the total water in these pigments by combustion with plumbic chromate. This is apparently the only way of obtaining a correct result, as, on heating, part of the hydrogen is evolved as hydrocyanic acid, the oxygen of the water going to form ferric oxide, as stated. The nitrogen was determined as above described, and was found to differ within small limits. A weighed quantity was ignited, care being taken that the teiiiperature was sufficiently high to decompose the last traces of the blue, but not too high to render the oxide of iron difficult of solution. It is necessary to ensure the complete oxidation of the finely-divided iron, as otherwise it is very difficultly soluble in hydrochloric acid.In all cases on adding acid to the "ash," a marked effervescence showed that alkaline carbonate (from the alkali-containing ferrocyanide) was present. Pure blue leaves an '' ash" completely soluble in hydrochloric acid, but in the case of adul- terated Prussian blue or of diluted blues, barium sulphate or other insoluble matter must be filtered off and weighed before precipitating with ammonia. The iron in the (weighed) mixed oxides was determined by titration with potassium bichromate after reduction with stannous chloride, the alumina being of course estimated by difference.A portion of the filtrate was evaporated to dryness, and, after volatiliza- tion of the ainmonium salts, the alkaline salts were weighed, and the chlorine therein '. Loco citato.THE ANALYST. Water of combination, etc. . * Cyanogen . . . . . . . . . j- Iron ... ... . . Aluminium . . . . . . . . . Alkali metal. . , . . . . . . Alkaline sulphate . . . . . . * Nitrogen . . . . . . .. . t ‘ I Extra radicle ” iron . . . Intra-radicle iron (found) . . . Intra-radicle iron (calculated from nitrogen) ... ... 227 I. 16-38 39 -96 31-23 1 .95 8 -05 2-45 100~00 21 -52 1 7-65 13 -58 14.34 determined by titration with silver nitrate.I n another portion of the filtrate the sulphuric acid was determilzed, and from these data--via., the weight of the alkaline salts, chlorine, and sulphuric acid-we were enabled to show that in all cases the alkali-metal was almost entirely potassiurii or sodium, and in no case a mixture of the two metals. Another weighed quantity of the pigment was boiled for a few minutes with aqueous potash, and the resulting precipitate of oxide of iron weighed, after ensuring its complete conversion into Fe,O,. This, multiplied by 0.7, represents the ‘‘ extra radicle” iron. The following t,ables represent the coinposition of eight sainples of blue sold by makers of repute as genuine Prussian blue of good quality : CO3IPOSITION O F EIGHT SARIPLHS O F COM3IERCIAL “ P U l i E ” PItUSSIAN BLUE, 11.18-85 4 2-61 33.34 -54 4-66 .~ - .~ 100~00 22-94 19’69 13-65 15.30 I. 22’69 17-12 15.55 15-12 11. 22’91 23‘30 20.09 20-51 13.73 12’78 15.28 15.54 111. IV. V. V1. VII. V I l I . Moisture (lost at 100” C.) ... ... Water of combina- tion, etc. ... ... * Cyanogen ... ... j- Iron ... ... -4luminium .. , ... Alkali metal (Xa) ... Alkaline sulphate . . , 5 61 15-46 37-72 29-48 1’82 7-60 2.3 1 3-54 18-18 41‘10 32’16 -5 2 (K) 4-50 - 5-36 6.22 42-97 34.27 - (K) 7-72 3.46 5 -4 5 13.07 37’90 80.32 :3 ‘1 7 (K) 2-25 7-84 $4.53 3-08 10.64 7.92 ‘72 (K) 1.06 2 00 5.32 7-86 39-91 30.94 1.00 (K) 11.31 3.66 5.56 14-60 40.19 31.94 1.43 3-76 (Sa) 2.52 l(JO.00 21.49 16.21 100-00 100~00 22-13 18.99 100 .oo 23.14 19.05 100.00 20.18 16-04 190-00 -5.73 - 100-00 21-64 18-98 1 @O.OO 22.00 19-38 * Nitrogen ...... i- “Extra radicle’’ iron CORIPOSITION OF D R Y MATTER O F ABOVE S A M P L E S ( D R I E D .4T 100” c.). VI. I VII. I VIII. IV. V. 12.03 41.i; 31.30 2-90 4-16 7 -84 111. 6-58 45-41 36’21 8-18 3 6 4 - 13-87 40.06 32‘05 3 -35 2-38 8’29 8 3 5 42.14 32.67 1 -03 11-94 3 .87 15.46 42.55 33-82 1-52 2-87 3 -98 17-95 43-29 33-09 1-61 0-81 3.25 100-00 24’45 20’02 16’19 16.30 100-00 21.57 16.95 15-10 14-38 In looking at these figures, it will be seen that the percentages of iron and nitrogen do not vary greatly. The amounts of iron in the iron-cyanogen radicle, calculated from the percentage of nitrogen and from the difference between total and extra radicle iron, do not agree closely in all cases. We can offer no explanation of this.Taking the figures for the dry matter, the greatest difference in iron was in the cases of 111. and I., 4.98 per cent., and the greatest difference in nitrogen- between 111. a,nd I. again-2-93-differences which, calculated on the mean per- centages, are under 16 and under 13 per cent. of the totals respectively. I t will also be noticed that seven out of the eight samples contained aluminium. This is not to be considered as an adulterant, as alum is often added to the sulphate228 THE ANALYST. of iron used in precipitating the ferrocyanide without any idea of adulteration, but probably to cause the precipitate to settle better and to dry in a more coherent manner. It probably exists as aluminium ferroeyanide, a compound described by Tissier (Comp.reud., xlv., 232). This has no colouring properties, and, of course, the iron and nitrogen in the iron cyanogen radicle of this compound appear in the total iron and nitrogen as determined. If merely the ammonia 2wecipitute is taken and called ‘‘ ferric oxide,” the alumina is included as oxide of iron. We think that the addition of alum to the precipitating tank is unnecessary, especially as the best sample of blue we have examined contained none. Alkali metal, as iron-akali-cyanide, is present in every case. This agrees with the statement of Williamson and others that a blue cannot be obtained free from alkali-metal, except when hydroferrocyanic acid is substituted for alkaline ferro- cyanide in the manufacture (the alkaline sulphate present in each case is, of course, due to imperfect washing).I n fact, Prussian blue, if by that term we mean ferric ferrocyanide Fe-(CN),,, is not known to the commercial world. Coinrnercial Prussian blue is a mixture of Williamson’s blue with other iron-alkali cyanides, and often with aluminium-iron cyanides, altogether a most complex substance. An examination of the watery extract showed, in the cases of I. and VI., the presence of alkaline ferrocyanide. This is rather remarkable, as with the quantities usually taken the iron is in excess. We believe that these came from the same maker, but are sure, however, that both samples were sent out as boni-fide Prussian blue. The following notes on appearance of samples in lump and in powder, depth of colour, etc., may be of interest : Condition.I. Lumps, conchoidal fracture ... 11. Lumps ... ... 111. Lumps, conchoidal fracture ... IV. Lumps, conchoidal fracture ... VI. Lumps, conchoidai fracture ... V. Paste ... VII. Powder ... ... VIII. Lumps . _ _ ... Appearance. blackish copper glance, but not a blue, pale-looking, no coppery glance bright sheen ... ... ... rich blue, bright coppery sheen ... ditto, but not as good as 111. dried, with coppery lustre . . . ... ... blackish coppery glance, but not a bright sheen ... ... ... light blue, no coppery glance ... Powdered Samples. Classification in order of Depth of Colour. 2 G 4 3 - 1 7 5 2 u ali ty )f Tint. The mean percentage of iron in the dry matter of these eight samples was 3296. The mean This would give 3.03 as the factor for calculating iron to Prussian blue.THE ANALYST.In Dry Matter. N x 4.4 ... ... ... Fe x 3.03 ... .. ... 229 r. 11. 111. IV. v. VI. VII. VIII. 94-69 100.93 107.58 94.91 98.95 99.83 1OOm80 102.52 94-63 101.02 109.71 9'7.11 94.84 98.99 102.47 100.26 ~~~~~~~~ Of course, these are merely empirical factors, based on the examination of com- mercial samples of varying composition, and do not possess the accuracy of a factor to reduce one definite chemical compound to terms of another, but, at any rate, taken together they are of sufficient accuracy for calculating the percentage of Prussian blue in pigments containing that colonr. The nitrogen alone in most cases would be sufficient. We feel justified in stating that blue sold as Prussian blue should contain a t least 20 per cent.of nitrogen and 30 per cent. of iron calculated on the dry matter, and, after burning, should be entirely dissolved by hydrochloric acid. A '' dry " blue should certainly contain under 7 per cent. moisture. Another important point is that the sulphuric acid used in the Kjeldahl nitrogen determination should not be blackened. Pure cyanides dissolve in sulphuric acid without any charring, which, if it occurs, indicates the presence of organic adulterants. A case of this kind will be noted later. The watery extract should not indicate the presence of excess of slkaline ferro- cyanide. We consider that this is evidence of defective manufacture, although it cannot, of course, be regarded as adulteration. Still it-containing as it does iron and nitrogen-is partly included in the analysis as Prussian blue, whereas it has no pigmentary value, and its solubility in water does not add to the weathering qualities of the colour.COMMERCIAL PIGMENTS. The following cases of examination of an adulterated sample of Prussian blue (of German manufacture) and of some colours ground in oil possess soine interest. A rather pale-looking sample of dry powdered Prussian blue, which we were told was a cheap sample of German manufacture, gave only 14.35 per cent. of nitrogen. This at once aroused our suspicions, especially as, instead of becoming colourless in a few minutes on boiling with sulphuric acid, a decided blackening occurred. This clearly pointed to the presence of an organic adulterant;. On further examination, the following figures were obtained : Moisture ...... ... ... ... ... ... 11-35 Water of combination, organic adulterant, alkalies, etc. 37.76 Iron ... ... ... . . . ... ... ... 22.67 Aluminium ... ... . . . . . . . . . ... ... 1-14 Insoluble ... ... ... ... ... -43 ... Cyanogen* ... ... ... ... ... 26.65 100~00 Per cent. nitrogen x 4.4 ... ... ... ... ... 63.10 ,, iron x 3.03 ... ... ... ... ... 68-69 * Nitrogen ... ... ... ... ... 14.35230 THE ANALYST. This sample not only contained more water than is usual, but was clearly deficient in blue. We think that one is justified in stating that it did not contain more than 70 per cent. of Prussian blue. Brunswick green is a Prussian blue -lead chromate green mixed with barium sulphate ( i ‘ barytes ”). A portion from which the oil had been extracted was gently ignited and treated with hydrochloric acid, the solution of lead, iron, and chromium chlorides filtered from the barium sulphate, and the lead precipitated with sulphuric acid. In the filtrate from the lead sulphate the ‘ I ammonia precipitate ” containing the mixed oxides of iron and chromium was determined, the oxides were separated by fusion with alkaline carbonate and nitrate and extraction of the fused mass with water, Fe,O, being insoluble.Direct titration with K,Cr,07 was difficult, as the colour of the chromium inasked the reduction of the iron with stannous chloride : Two samples of Brunswick green ground in oil were examined. Light -green. Dark-green. Oil ... ... ... ... ... 11.27 ... 13-55 Chrome yellow ... ... 4.98 ... 7.10 Barium sulphate ...... ... ... 82.74 ... 75.14 Prussian bi;lk (Fe x 3.03) ... ... 1.01 ... 4.21 ... ... 100*00 100~00 The nitrogen was determined in each of these samples in about 5 grammes, and was -19 and -72 respectively, which, using 4.4 as a multiplier, would give -83 and 3.16 per cent. of Prussian blue in these ssmples. We are much more inclined to rely on a direct determination of this kind than on the determination of the iron by the method described. The method of analysis for these colours described by Hurst, and due to Brown (Chenh. News, December 31, 1886), seems still more unsatisfactory. The green is treated with hydrochloric acid to dissolve out the lead chromate (of course, as lead and chromium chlorides). The residue, barytes and Prussian blue, is ignited to decompose the blue, weighed, treated with aqua regia to dissolve out oxide of iron (and alkalies), and the insoluble portion again weighed. The difference is to be taken as oxide of iron and multiplied by 2.212 to give Prussian blue. The lead and chromium are, of course, determined in the original filtrate. I n the case of colours ground in oil, the Kjeldahl method is particularly applicable for the determination of nitrogen, and, as now it is very general to sell these colours in oil instead of dry, we think that the nitrogen determination will be found a very convenient method of ascertaining (approximately) the amount of Prussian blue in the many diluted colours made from it. Adulteration with nitrogenous matter is not probable, and so is not likely to lead to erroneous results. We intend t o continue the examination of all samples which come under our notice by this niethod, which is rapid (we have obtained a result in under one hour), and at once points to any serious sophistication. I n conclusion, we have to mention that this work was done in the laboratory of the London County Council in the course of our regular official duties. We also desire to express our thanks to Messrs. S. Lambert and Go., Sir William Rose and Go., and Messrs. Colthurst and Harding, for providing us with samples and information respecting their manufacture.

ISSN:0003-2654    

DOI:10.1039/AN8962100225

出版商:RSC    

年代:1896

数据来源: RSC

摘要:

THE ANALYST. 231 ABSTRACTS OF PAPERS PUBLISHED IN OTHER JOURNALS. FOOD ANALYSIS. Iodine Number and Refractive Index of Cocoa Butter. A . Strohl. (Zed. aiial. Chern., 1896, xxxv., 166-169.)-1n order to determine what were the actual limits in these constants, the author examined samples of all the known trade varieties. The iodine number was determined by Hiibl's method, an excess of iodine (over 60 per cent.) always being allowed. The results showed that the iodine number of pure cocoa butter, whatever its source or method of preparation, varied between 32-8 and 41.7. The refractive index determined at 40" C. varied between 1.4565 and 1.4578, these numbers corresponding to 46 and 47.8 on the scale of the Zeiss butyro-refrac- tomer. A certain parallelism was noted between the iodine number and refractive index, a cocoa butter with a low Hiibl number having also less refractive power.(Cj. Hehner, ANALYST, xx., 136.) C. A. M. Testing for Darnel in Flour. A. Deros. (Ann. d e C h h . Anulyt., i. [lo], 190, 191.)-As the colour-reaction with sulphuric acid, described by Wagner and Ginsberg ( A m . de C'lzinz. Anal:yt., i. [8]), is due to saponin, it cannot be considered as characteristic of darnel alone. Corn-cockle, which is frequently present in wheat, barley, and other flours, also contains saponin. Neither do the properties of the fat extracted by ether, nor the polygonal form of the starch grains (when in small pro- portion), considered separately, afford convincing proof of the presence of darnel ; but when these characteristics are cumulative a positive decision may be given.c. s. Chemical Detection of Horseflesh. Courlay and Coremons. (Ann. d e ??zed. veterin. Through Zcit. Nahy. Untei*such, 1896, 173, 174.)-The method here described is based on the same principle as that of Brautigstm and Edelmann (ANALYST, xx., 95 and 252), but is claimed to be more easy of application. About 50 grammes of the substance to be examined in as fresh a state as possible are finely divided and boiled from fifteen to thirty minutes with 200 C.C. of water. After cooling, the broth is filtered, and then tested with a few drops of iodine solution, prepared by dissolving 2 parts of iodine and 4 parts of potassium iodide in 100 parts of water. A brown coloration, disappearing on warming to 80" C, and reappearing on cooling, shows the presence of horseflesh.In the case of the addition of flour (e.g., in sausage) the blue d o u r may mask the glycogen reaction. This is obviated by adding two or three times the volume of concentrated acetic acid to the broth, filtering, and again testing the filtrate wit;h the iodine solution. KO reaction was obtained by the authors with the flesh of cattle, calves, pigs, dogs or cats, but this observation does not apply to the flesh of the fetus of any of these animals. C. A. M. Horse-flesh and Horse-fat. R. Fruhling. (Zeit., ayzg. C'hem., 1896, 352, 353.) -Not being able to obtain conclusive results with Edelmann's glycogen reaction, the232 THE ANALYST. author has attempted to use the iodine number of the fat as a means of detection.Fat taken from different parts of a freshly-killed horse gave the following values : Specific gravity at 500 ... at 17.5" ... 1odi;e nuLber (Hiibl) . . , Kottstorfer number , . . . ... Iodine number of fatty acids Melting-point of fatty acids ... Solidification - point of fatty acids . . . ... ... ... Back. 0-8963 0.9159 79.9 182.8 81.4 52-53" 45-43" Heart. 0-8948 0.9167 77-4 184.7 78.3 40-41" 34-32" Kidney. 0.8987 0.9212 82% 187.6 84-0 53-54" 48-47" Mean. 0.8966 0-9180 80 .o 185.0 81.6 48.8" The iodine numbers of the fat extracted from three specimens of sausage con taining horseflesh were as follows : Iodine number. Sausage made from pure horseflesh ... ... ... 72-5 ,, consisting of horseflesh, with 15 per cent. of pork 62.3 # ¶ 9 9 9 , , I 50 ,, ?, 57.2 Since lard has an iodine number of from 56.9 to 63.8 (according to Benedikt, pork), it is obvious that this method would lead to no certain conclusion.C. A. M. The Estimation of Caffeine. M. Georges. (Jozw. Phnwn. ct Chim., 1896, xvi., 58, 59.)-The sample (tea, etc.) is finely powdered, and 5 grammes mixed with fine sand, and lixiviated with a 1 per cent. aqueous solution of salicylate of soda until the caffeine is completely extracted. The liquid is concentrated to about 50 C.C. on the water-bath, filtered, and the filter washed with the salicylate solution. The caffeine is then extracted from the solution with chloroform, and the latter evaporated. The caffeine is perfectly white and free from impurities. In a sample of black tea analysed by this method 3-93 per cent.of caffeine was obtained. No comparative figures of the results obtained by other methods are given. Sodium benzoate or sodium cinnamate may be substituted for sodium salicylate, but caffeine is not quite as soluble in their aqueous solutions as in that of sodium salicylate. C. A. &I. The Estimation of Caffeine in Tea. A. Petit and P. Terrat. (,Jnn. de C'liint. AnaLgt., i. [la], 228.)-The researches made by the authors show that, instead of magnesia or lime being necessary to liberate the caffeine from its state of combination in the tea, as supposed by Comniaille, before it can be extracted by chloroform, water alone is sufficient to effect this result; and, further, that it is necessary to perform the extraction on moist tea, since the caffeine resumes its insoluble condition during the process of drying.Referring to the method, proposed by Paul, of extrac- tion by alcohol, they find that this solvent requires dilution to 60" or 80" strength, 98" alcohol only extracting 0.88 per cent. of caffeine from a tea containing 2.50 per cent. The maximum was obtained by means of the Commaille method performed with a moist mixtzm, and by the Petit and Legrip method, which latter consists in treating 25 grammes of powdered tea with three times its weight of boiling-water for a quarter of an hour with frequent shaking; then evaporating on the water-bath until the tea just merely moistens the fingers when pressed, whereupon the chloro-THE ANALYST. 233 form extraction is performed, and continued until the residue, after being taken up by boiling water and filtered, no longer gives any precipitate or turbidity with a solution of tannin. The chloroform being then distilled off, the extract is taken up by boiling water, filtered through a moist filter, carefully washed, evaporated on the water-bath, and weighed.If the crude caffeine contains chlorophyll, it should be treated by Grandval and Lajoux’s method--viz., dissolving in 15 C.C. of 10 per cent. sulphuric acid in the cold, filtering, neutralizing the acid by ammonia, and evaporating to complete dryness, taking up the residue with chloroform and evaporating this solution at a low temperature. I n all the methods tested, wherein the extraction by chloroforni or alcohol (except dilute alcohol) was performed on dry tea, whether with or without the presence of lime or magnesia, the amount of caffeine obtained was far below the truth, and although the method of Grandval and La joux-who substituted ammonia for the other alkaline bases mentioned, and employed ether in the formation of the mixture-gave nearly as good results as that of Petit and Legrip, described above, the latter is considered to be preferable as being simpler.The alcohol methods take more time, and the purification of the caffeine is more difficult than by the chloro- form method. c. s. On the Estimation of the essential Oil of Mustard in Feeding Stuffs. 1\6. Passon. (Zeit. ccngewniz. Chenz., 1896, 422, 423.)-The process here described depends on the fact that mustard-oil is soluble in glacial acetic acid, and, as the author’s experi- ments prove, may be quantitatively determined in such solution by Kjeldahl’s method.The apparatus required consists of a flask, u, fitted with a safety tube and a tube delivering into a second flask, b , frorn which passes a tube into a small beaker, c , containing 20 C.C. of sulphuric acid, to absorb any mustard-oil escaping from b. The flask b contains from 25 to 75 C.C. of glacial acetic acid and a little zinc and iron dust to partially reduce the mustard-oil to allylaniine. About 25 gramines of the oil-cake are mixed with 300 C.C. of water and 0.5 gramme of tartaric acid in flask a, and slowly distilled for about two hours. The contents of b and c are then mixed in a Kjeldahl flask, and the nitrogen determined in the usual way. One C.C.of decinormal alkali corresponds to 0.0099 gramme of ally1 iso-thiocyanate (C,H,NCS). I n a sample of rape-seed cake the author found (1) 0.205 per cent. and (2) 0.219 per cent. of mustard-oil. In the second case the apparatus was allowed to stand for twenty-four hours, and the zinc and iron dust not added until imm-ediately before the distillation. C. A. &I. Application of the Rontgen Rays to the Analysis of Vegetable Substances. F. Ranwez. (RzdZ. d e ?Ass. bclge, 1896, x., 44-48.)-The author describes how, by means of photographs taken with the Rontgen rays, adulteration of vegetable sub- stances with mineral matter may be readily detected. So far his experiments have been confined to the examination of samples of saffron, some of which were pure, and others adulterated with barium sulphate.Detection of Egg-yolk in Pastry, etc. C. A. M. E. Spaeth. (Zeit. Nahr. Untersz~h, 1896, x., 171-173.)-1n order to determine whether flour products have been coloured234 THE ANALYST. with yolk of egg or with other substances such as saffron or picric acid, the author examines the fat. According to Gobley and others, egg-yolk has the following per- centage composition : Water ... ... ... 5.18 ; Cholesterin ... ... 0.4 Vitellin ... ... 15% Glycerin-phosphoric acid 1.2 Nuclein ... ... 1.5 1 Lecithin ... ... 7-2 ... ... 20-3 Colouring matters ... 0.5 €'dft in-\. Cerebrin -.. ... 0.3 Stearin J Salts ... ... ... 1.0 Of these constituents, triolein, tristearin, tripalmitin, cholesterin, and glycerin- I n the subjoined table the characteristics of the fat are given, together with those phosphoric acid are soluble in ether.of the fat from wheat flour : Egg-yolk Fat. Wheat-meal Fat. Specific gravity at 100" C. (water at 15"=1) ... ... ... ... ... 0.881 ... 0.9068 Melting-point of fatty acids ... ... 36" ... 34" Saponification number (mean of 3 results) 184-43 ... 166.5 Iodine number (mean of 3 results) ... 68.48 ... 101.5 Iodine number of fatty acids ... ... 72-6 ... Reichert-Meissl value ... ... ... 0-66 ... 2.8 Refractive index at 25" C. ... ... 1-4713 ... 1.4851 scale ... ... ... ... ... 68.5 ... 92.0 - ,, ,, on Zeiss refractometer Although a quantitative determination is not possible, yet by means of the fat an approximate judgment as to the amount of egg-yolk present can be formed. Flour products with only a slight addition of yellow of egg invariably give more ether extract than those derived from wheat meal and coloured.For the examination from 100 to 200 gramrnes of the flour products are ground as finely as possible and extracted with ether in a Soxhlet's extractor. The ether is evaporated, and the residue taken up with petroleum spirit of low boiling-point. After standing several hours until no more solid matter separates, the liquid is filtered, the petroleum spirit evaporated, the residue dried in the water-oven, and its iodine number and refractive index determined. When the iodine number exceeds 98, and the phosphoric acid in the fat is below 0,005 per cent., there cannot be more than traces of egg-yolk. Of course, it must not be lost sight of that wheat-flour itself also contains phosphoric acid.C. A. 31. Estimation of Free Tartaric Acid in Wines. L. Magnier de la Source. ( A i m . c7e C'him. ,4izciZyt., i. [ll], 205.)-In order to ascertain whether the action of potassium acetate in preventing the precipitation of potassium bitartrate in a mixture of ether and alcohol is transitory or permanent, 60 C.C. of a wine, to which 13.5 grammes tartaric acid per litre had been added, were divided into three equal parts, to each of which were added 100 C.C. of ether-alcohol mixture, and, in addition to this, to the first (n), 5 C.C. of distilled water, to the second ( b ) , 5 C.C. of concentrated potassium acetate, and to the third (c), 5 C.C. of a concentrated solution of potassium bromide. After stanzing for two days at a temperature approcchiag zero, the tartaricTHE ANALYST.235 acid was estimated, and found to be, in a, 6.96 grammes, in b, 11.60, and in c, 19-20 grammes. The amount retained in solution by the potassium acetate in b was there- fore 18-24 grammes, practically the whole of which can be precipitated in 48 hours by means of pot,assium bromide. Of the 19-20 grammes found in c , 2.50 grammes were present originally in the wine, leaving 16-70 grammes corresponding to 13.30 gramnies of the 13.50 grammes of tartaric acid added, whereas only about 50 per cent. of the tartaric acid is precipitated from h. If the filtrate from b be united with the washings and an excess of acetic acid added to one moiety, crystals of tartar will develop in the course of 24 hours, whilst the other moiety will not show any trace of deposit.These results show the defects of potassiurn acetate as a precipitant of tartaric acid, and at the same time demonstrate the exactness of the results obtained by potassium bromide (or chloride). c. s. Composition of Goose-fat. J. Rozsenyi. (Bc7'. Cheiiz. Imt., Buda Pest ; through Zeit. ang. CILCIIL., 1896, 364.)-The following constants were obtained with four samples of goose-fat rendered by the author : Specific gravity (15") . . . ... 3Ielting-point Jcommenced Solidification-point ... ... Melting -point, fatty acids (cam- menced) ... ... ... ... Melting-point, fatty acids (ended) Iodine number . , . ... ... Reichert-Meissl number . . . ... Kottstorfer number ... ... Hehner number .. ... ... Refractive index at 40" (Zeiss) . . . ... ... (ended ... 1. 2. 0.9229 0.9258 26.6" 30.0" 27.5" 31.4" 18.4" 18.3" c 1 '3 .I, 9" 35.3" 36.6" 37.5" 58.7 62.8 0.3 0.3 193.0 191-2 94.5 95.3 50.5 30.0 39.0" 40.2" 66-4 0-3 191.6 95.1 50.5 4. 0.9300 29.5" 31.0" 1S.l" - 62 -5 0.2 193-0 50.5 C. A. M. - The Amount of Copper in Vegetable Produce. V. Vedrodi. (Chem. Zeit., xx., 1896, 399.)-Considerable differences in opinion have arisen in the past between the author and other observers regarding this question, and Lehmann in particular has recorded a number of figures showing only about the one-hundredth part of copper found by the present investigator. Lehmann's process was a colorimetric one, the red colour produced by ferrocyanides being employed to determine 0.02 to 0.2 milli- gramme of copper, and the blue yielded by ammonia when as much as 0.2 to 1 milli- gramme was present.The samples were burnt in a porcelain crucible after addition of sulphuric acid, the ash dissolved in acid, the solution evaporated, and the ammonium salts filtered off. Vedrodi has always objected to the use of the acid, and prefers to incinerate the substance alone in a muffle, dissolve the ash in hydrochloric acid, pass sulphuretted hydrogen, filter, wash and ignite the precipitate, again dissolve and throw it down as before, finally weighing and calculating it into copper oxide. Finding that he originally made a slight mistake in the allowance for the weight of236 THE ANALYST. the filter ash, he has checked the process on minute quantities of pure metal, and has made a number of fresh determinations, the results of which are recorded in the table.He remarks that the form in which the copper exists in these substances is unknown ; and he appears to doubt the toxicity of some of the organic salts, pointing out that in many places pickled gherkins, etc., are habitually coloured with copper, and sold without any complaints arising about evil effects produced by their con- sumption. MILLIGRAMMES OF Cu PER KILO. Winter wheat Summer wheat Rye _.. Oats ... Buck wheat Haricots ... Lentils . . . Peas ... Soja beans Lupins ... Must ard-seed Capsicum.. . Barley ... 1894. -c--, Min. Max. ... ... 80 ... 710 ... ... ... 190 ... 630 ... ... ... 60 ... 90 ... ... 80 ... 120 ... ... . . . 40 ... 190 ... ... 160 ... 640 ... . . . ... 160 ... 320 ... . . . . . . 120 ... 150 ... ... .. GO ... 100 ... ... ... 70 ... 100 ... ... ... 80 ... 190 . . . 70 ... 130 ... ... ... 790 ... 1350 ... ... ... 1895. 200 190 10 10 40 150 110 110 60 70 70 GO 230 Mai . ... 680 ... 230 ... 30 ... 70 ... 200 ... 160 ... 150 ... 150 ... 110 ... 80 .. 290 ... 70 ... 400 F. H. L. Notes on Cubebs. K. Peinemann. ( A ~ c h . Pharwz., 1896, ccxxxiv., 241 ; through Chena. Z&. Rep., 1896, 162.)-Cubebin exists chiefly in the fruit of the cubeb plant, but it may also be found in the stalks, and to a lesser extent in other portions of the plant. It is present in the pericarp as well as in the perisperm, as is also the case with piperine in the ordinary pepper. With the exception of Lowong pepper, those peppers in which cubebin or its derivatives are to be detected never contain any alkaloid (piperine), and the converse of the statement also holds good. With strong sulphuric acid it gives a fine purplish violet colour. Dissolved in chloroform, it rotates light to the left, pseudo-cubebin to the right. Cubebin contains C,,H,,O, ; it crystallizes in needles, melting at 125" C. F. H, L.

ISSN:0003-2654    

DOI:10.1039/AN8962100231

出版商:RSC    

年代:1896

数据来源: RSC

摘要:

236 THE ANALYST. TOX I CO LOG ICA L ANALYSIS. Purification of Arsenical Sulphuretted Hydrogen by Iodine. Schlagden- hauffen. (Bzdletin Commercial, April, 1895 (?), p. 178 ; through Ann. de Chim. AnaZyt., i. [Ill, 209.)-Jacobson and Brunn (BericlLte, xx., p. 1099; xxi., p. 2546) remore the arseniuretted hydrogen present in sulphuretted hydrogen (prepared from commercial hydrochloric acid and sulphide of iron) by passing the gas over iodine, the iodide of arsenic formed collecting in the U-tube of the apparatus. If, however, the sulphuretted hydrogen is liberated too quickly, the action of the iodine is incomplete,THE ANALYST. 237 and besides, the investigations of Skraup, confirmed by those of the author, show that the arsenic is not all liberated in the early stages of the primary reaction, as was asserted by Otto, but continues to come oft’ throughout, wherefore it is necessary to pass the ziyhoZe of the gas-and not merely the first portions-over the iodine.The hydriodic acid formed simultaneously with the iodide of arsenic must be removed from the sulphuretted hydrogen by absorption in water, otherwise, in cases where the sulphuretted hydrogen is allowed to react on metals or alloys contaminated with arsenic, the iodine will be present in the ammoniacal solution submitted to the Marsh test, and, being converted into iodic acid by the action of the nitric acid employed in the test, will monopolise the first portions of the nascent hydrogen formed, and so retard the deposition of the arsenical ring, thereby leading to the conclusion-unless the duration of the test is prolonged-that there is no arsenic present. The author prefers to employ so-called pure hydrochloric acid instead of the commercial acid which often contains considerable quantities of arsenic. c.s. Detection of Mercury in Cases of Poisoning. D. Vitali. (Chenz. Zeit., xx., 1896, 517.)-The organic portion of the substance under examination is destroyed according to Fresenius and Babo’s process, and the filtered liquid evaporated as far as possible without causing the mercuric chloride to crystallize out. I t is then treated with a stream of sulphuretted hydrogen for several hours, being kept warm all the time. The sulphide is washed by decantation, dissolved on the water-bath in aqua regia, and the chlorine and nitric acid removed by evaporation with hydrochloric acid. The solution is introduced into a porcelain basin, in which are placed a piece of goldfoil and an iron nail. In an hour’s time the whole of the mercury will probably be deposited partly on the gold and partly on the nail. These are taken out, washed and dried, put into a small test-tube, and ignited gently. The mercury yields the characteristic gray sublimate, which may be identified with greater cer- tainty by adding a minute crystal of iodine, and heating gently once more; if an excess of iodine is used, the yellow or red colour of the iodide may be masked, but on warming in a current of air it will be seen distinctly. The reaction detects 0.01 milligranime of mercury. The iron and gold may also be heated on the water-bath in a basin covered with another, the outside of which is moistened with gold chloride, when the gold is reduced to the metallic state. F. H. L.

ISSN:0003-2654    

DOI:10.1039/AN8962100236

出版商:RSC    

年代:1896

数据来源: RSC

摘要:

THE ANALYST. 237 ORGANIC ANALYSIS. Action of Nascent Bromine on Naphthols and Naphthylamine. W. Vaubel. (Zait. anal. Chem., 1896, xxxv., 164-166.)-With a view to subsequent application to analytical processes, the author has studied the action of nascent bromine liberated by adding hydrochloric acid to a solution of potassium bromide and bromate, on naphthol and naphthylamine derivatives dissolved in acetic acid. u-Naphthol.-Absorbs two atoms of bromine rapidly. Subsequently a further238 THE ANALYST. absorption slowly occurs, which renders the end reaction obscure. bromination method for estimating the amount of this substance is not practicable. naphthol ethyl ester. Melting-point, 48" C. End reaction distinct. Consequently a Ethyl Xster of a-Naphthol.-Absorbs one atom of bromine, forming the brom- /3-NuphthoZ.--Absorbs one atom.Methyl Ester of P-NaphthoL--Absorbs one atom. End reaction clear. The bromide separates in white shining needles. Ethyl E'ster of P-Naphthol.-Behaves like the methyl compound, but the absorption is slower. a-h7aphthylamine.-In the main absorbs two atoms of bromine. Owing to the simultaneous formation of small quantities of oxidation products, bromination cannot be employed as a method of estimation. a-dcetizcL~lzt~iylaiiLi?zc.-Absorhs one atom. End reaction well marked. The mono-bromide melts at 190" C. P-Na~~lLthyZamiize.-Rapidly absorbs two atoms. End reaction not clear, owing to formation of oxidation products. P-Acetizn~hthylnnziize.-One atom absorbed. End reaction clear. Compound iiielts at 140" C.End reaction well marked. Melting-point, 85" C. The mono-bromide separates in plates, melting at 54" C. C. A. M. Adulteration of Rape-seed. Pajot. (Aim. de Chim. Aizalyt., i. [lo], 187-190). -The author having been called upon to report on the percentages of oil in certain oleaginous seeds, discovered a somewhat extensive adulteration of German rape-seed by two varieties of mustard, viz., charlock and Brassica juncea, coloured to resemble rape. The blue colouring matter is easily extracted by cold water in about fifteen to thirty minutes, or much more readily by the action of an acid, in which case a rose- coloured solution, turning blue on the addition of an alkali, is obtained. The test may also be performed by placing the seed on a white filter-paper moistened with acid or water, the colour appearing at the points of contact.The aqueous extract is found to contain salts of iron and sulphates, and has an alkaline reaction, whereas pure rape-seed extract is acid, and contains neither iron nor sulphates. The colouring matter is believed to be a product analogous to litmus, and, being applied when the seeds are newly harvested, is fixed and modified by the integument of the seed. A considerable profit must be derived from the fraud, since the charlock is only worth about three shillings per cwt. and the Brassica juncea-seeds about six shillings, whereas rape-seed averages about nine shillings per cwt. The yield of oil is from 10 to 15 per cent. less than from pure rape-seed, and its quality inferior. I n addition to this, the oil-cake is rendered unfit for cattle food, and is only suitable for manure.C. S. Estimation of Glucose. Causse. (JOZLT~. de Phurmcicie, 1896, p. 433 ; through A m . de Chim. Analyt., i. [loll 192.)-The author criticises the method proposed by Gerrard (Phar?mcez&xd J o k z . , April, 1895), and considers that his own process (Jozmz. dc Phar?n., xix., 171) is preferable, since the potassium ferrocyanide employed to prevent the precipitation of the cuprous oxide has no action on Fehling's solution.THE ANALYST. 239 Good results are obtained by the author's method in the detection of sugar in urine and blood. c. s. - ~ _ _ -. On the Distillation of the Lower Acids of the Fatty Series in Presence of Water. E. Sorel. ( A m . c7e Chiin. Amdyt., i.[13], 245.j-The difference between the proportions found by Duclaux and Muntz respectively in their researches on this subject, is due to the conditions under which the experiments were made-ie., to the modifications produced by condensation in the composition of the distillates prepared in retorts exposed to radiation. To ensure regularity of results, the retorts should be surrounded by a metallic jacket, heated by a ring-burner to a temperature exceeding the boiling-point of the still contents before distillation is commenced ; and, further- more, the neck of the retort should be bent so as to prevent the return of any con- densation products to the still. The basis of the calculation of the ratio of the acids in the fractions of distillate to the amounts in the liquid under examination is determined by distilling 550 C.C.of a liquid of known composition, dividing the distillate into nine portions of 50 c.c.-- disregarding the first 50 C.C. as containing the products passing of'f below the boiling- point-and leaving the tenth portion of the liquid in the still. The purity of the acids is checked by the preparation of their compounds with barium. From the table prepared by the author, it is deduced that the Duclaux method -with different co-efficients-is applicable to formic acid in the presence of acetic acid, provided they do not together constitute more than 10 per cent. of the original liquid; but in the case of the higher acids t;he method is not conclusive, particularly as the exact conditions of Duclaux' and Illuntz' methods of working are unknown c.s. The Valuation of Glue. C. Stelling. (Chem. Zeit., xx., 1896, 461.)--As the value of glue depends simply on its adhesive power, the only reasonable method of testing it is to attempt to determine the amount of substances other than gelatin (" non-glues ") it contains. These bodies react with tannin in B manner similar to glue itself ; while the quantity of water the sample will absorb depends chiefly on the process and temperature of manufacture. The following method gives comparative results, and at least indicates the source of the product, whence some idea can be obtained as to its probable value. 15 grammes are soaked over night in 60 C.C. of water in a 250 C.C. flask. Next day the jelly is dissolved on the water-bath, and the loss by evaporation made up.The flask is then filled with 96 per cent. spirit, and the whole thoroughly shaken. After standing 6 hours, 25 or 50 C.C. of the liquid are filtered off, the residue consisting of the non-glues dried at 100" and weighed. Five different sorts of gelatin gave 2.53 to 4-53 (mean 3.39) per cent. of non-glues; five glues gave 2.0 to 4-70 (mean 3.49) ; three leather glues 4.30 to 7.60 (5.73) ; four bone glues from acid bones 9-24 to 11-84 (10.33) ; three other specimens of the same 13.16 to 16.78 (15.15); seventeen pressed glues from neutral bone powder 14-30 to 32.10 (20.66) ; one of whale glue 23.0 ; and two of a material used for the clarification of wine 33.20 and 59-30 per cent. of non-glues respectively. F. H. L.240 THE ANALYST.Estimation of the ‘‘ Weighting ” Materials in Silk. H. Silbermann. (Chem Zcit., xx., 1896, 478.)-In the case of white or pale coloured silks, whether in skein or woven, the soluble weighting matters are extracted by boiling with water. The sugars are estimated after inversion by Fehling’s solution, and the magnesium salts and the sodium sulphate by sodium phosphate and barium chloride respectively. The residual silk is burnt, and the ash dissolved in hot, fairly strong hydrochloric acid. If tin only be present, as may be told by the behaviour of the acid liquid with sulphuretted hydrogen, the weight of the ash may be multiplied by 1.13 to give the amount of stannic acid originally in the fibre. To calculate how much weighting substance (E) is present in 100 parts of the material, the following formula may be employed : where CL = the original weight of the sample, b = the weight after extraction with water, p = the tin oxide ash, and A the loss in weight during boiling of the fibre itself, being 20 to 25 for ( h i t (Boiled silk), 5 to 9 fur Souple (Softened silk), and 0 to 2 for Ecru :- The tin may also be determined by igniting the sample in a porcelain crucible with an equal weight of caustic soda and a little potassium nitrate.The melt is extracted with water, the solution acidified and the tin thrown down by H,S. Another sample is boiled in dilute HCl to dissolve the lakes of tin and other metals; if the solution, treated with sodium acetate in excess and a ferric salt, yields a blue black precipitate, the estimation of the individual ingredients is conducted as follows. I n the aqueous extract the tannins are precipitated with gelatin; the sugars determined in the filtrate, and the tannins themselves in the deposit.The sample is boiled with dilute IICl and NaHO, and the tannins are precipitated from both liquids. Another portion of the material is burnt to ash, and if the latter is not completely soluble in acid, it contains barium sulphate or silica, or both, as may be ascertained in the usual manner. Dark silks may also contain iron and chromium salts, as is shown by the brown or green colour of the ash. This should be fused with NaHO and KNO,, the melt dissolved in water, which leaves the iron oxide behind, the solution acidified, and examined for the other metals.Black silks are boiled alternately for half an hour in NaHO (20 grammes per litre) and HC1 (250 grammes per litre) four times in each liquid, the fibre being thoroughly washed with hot water between each treatment, and finally cleaned by boiling in soap. If P = t h e original weight, and D the weight after boiling, the amount of weighting (E) is as follows :- (100 - A)(P - U). D E = I n this instance the value of A is 25 for Cuit, 8 for Souple, 0 for Ecru, and 10 for Fantasie. It is still possible for the fibre to retain metallic salts, and if this be the case it should be burnt to ash. Calling the weight of the latter A, the weight of theTHE ANALYST. 241 oxides in the fibre averages 1*25A, and therefore the true fibre in P is D - 1.25A. The weighting matters in 100 of the original material are (P - I> + 1-25A)(100 - A).D .- 1-258 E = (As an example, P = 80 grammes, D = 62, and A = 1-6 ; therefore E = 25 per cent. in the material ; while it contains 80 - 62 + 2 = 20 gramnies of weighting to 60 of true fibre, or 33 per cent, of foreign constituents,) An alternative process of estimating the true silk fibre in the material consists in a nitrogen determination. The nitrogenous substances, such as tannin, gelatin, and Prussian blue, are first rernoved by boiling in acid and alkaline carbonate, employing, in the case of Souple and Ecru, the ammonium salt instead of sodium carbonate. Thus treated, the fibre contains nothing but metallic oxides, and as the amount of nitrogen (17.6 per cent.) in silk is accurately known, the weight of the charge can be readily discovered.Either a combustion with soda-lime or Kjeldahl's process may be employed. The sample must be thoroughly dried, 2 grarnmes weighed off, and separated into distinct fibres with great care, in order to avoid the production of dust. With this object Persoz has suggested that it should be moistened in 1 : 3 HCl and then dried and powdered. If CL is the weight of the dry sample, and 1) the amount of nitrogen found, - - - 100 per cent. 1 0*176n(100 - a) E = [ . E being, as before, the amount of weighting substance in 100 of the material examined. With suitable precautions this process gives very accurate results. F. H. L. Saturation Temperatures and Critical Temperatures. Application to General Analysis.( B d c7e PAss. belge chim, 1896, ix., 359-375.)-1. &timation of Nitro-glyccri?z in Dymmite.-Several minor modi- fications in the process of determining the critical temperature of substances have been adopted (AXALPST, xx., 209, 257). When the temperature does not exceed 170" to 180" C., a glycerin-bath is preferable, and when the temperature is slightly lower, the platinum binding wire may then be replaced by an indiarubber band. Slightly larger tubes (diameter, 6 to 8 m.m. instead of 5 m.m.) are now used. The critical temperature of nitro-glycerin varies with the acidity, but with samples washed until completely neutral it is 46-5" to 46.7". In order to determine the relation in weight or volume between the two bodies giving the critical temperature of dissolu- tion, it is necessary to plot a curve of the temperatures of saturation.This is done by warming weighed quantities of the substances in small sealed tubes until clear, and noting the temperature at which they become turbid when allowed to cool in the bath. The saturation temperatures of nitro-glycerin in 90 per cent. alcohol are given in the following table, and are also plotted in a curve in the original paper : L. Crismer and J. Motteu.242 THE ANALYST. S i tro - glycerin. Grammes. 0.164 0.186 0.6053 0-6709 0-232 09403 0.2555 0.2595 0 2685 0.3135 04015 0.460 0.521 0,576 0.6425 Alcohol, 90 Per Cent. Grnmmes. 1 C.C. 3 C.C. 1 C.C. 9 2 Y ? ,? 9 7 Y , 2 , , 9 9 , 9 , 9 , ? I 9 9 0.818 0.8165 2.48'36 2.4928 0.817 0.813 0.820 0.8155 0.817 0.818 0.819 0.8175 0.811 0.8185 0.821 Nitro-glycerin. Per Cent.by Weight . 16.7 18.58 19.55 21-19 22.11 22.85 23.75 24.18 94.73 27-71 32-9 36 01 39-11 41.3 43 -9 Alcohol. Per Cent. by Weight. 53.3 81-45 80.45 78-81 77.9 77.15 76.25 75.89 75-27 79.29 67.1 64 60.9 58.7 56.1 remperature of Saturation. 0.4 7.8 11.4 15.4 18 19.8 21.6 22.6 23.9 29.6 36-9 40.5 43.1 44.75 46.5 '7.8 114 15.4 18 19.8 21-7 22.6 23.9 29% 36.9 40.5 43 -1 44.75 46.5 If x graiiinies of nitro-glycerin have been dissolved in 5 grammes of 90 per cent. alcohol, and the temperature found is 20.1", this would correspond to 23 per cent. of nitro-glycerin and 77 per cent. of alcohol, and the amount of nitro-glycerin can then be determined by proportion : 23 x 5 23 : 77=x : 5, or z = ---. 77 In estimating the amount of nitro-glycerin in dynamite, about 1 gramme of the latter is weighed into a small stoppered flask, together with 3 C.C. of 90 per cent.alcohol, about 2.5 grsmmes. The flask is then warmed for a few seconds at 30" to 35", well shaken, the liquid filtered through a dry filter, a small quantity sealed in a tube, and the saturation temperature determined as described above. The tabulated results obtained with dynamites prepared in the laboratory agree well with theory. Two sources of error are possible in this method : either the nitro-glycerin may not have the normal critical temperature, or the dynamite may contain a small amount of moisture. I n the former case a preliminary extraction of nitro-glycerin may be made with ether, the critical temperature determined after evaporation of the ether, and a fresh curve plotted on this basis.The latter error may be practically eliminated by increasing the amount of alcohol, but in such a manner that the tem- peratures of saturation still lie between 15" and 25". 2. Estimation of Water in Butter.-The following example illustrates this appli- cation of the process. The butter-fat gave a critical temperature 98", with 91 per cent. alcohol : Amount of butter taken .. ... . . . 0.7805 gramme. Water by evaporation ... ... . . . 0.1245 = 13.75 per cent, Alcohol containing 1.344 per cent. of water Critical temperature . . . ... ... ... 106.5 grammes. 1.219 gramme or 1-5 C.C. This temperature corresponds to an alcohol containing 10.5 per cent. of water, while the alcohol used contained 1-2028 gramme of alcohol and 0.01638 gramme of water. By the proportion 89-5 : 10.5 = 1.2028 : x the quantity of water in the alcohol broughtTHE ANALYST. 243 to 10-5 per cent. = 1.1411 grammes, of which 0.01638 grarnme was present in the original alcohol. Therefore the quantity of water derived from the butter = 0-1247 gramme, or 13.77 per cent. The method gives exact results when the butters are not salted, but when there is any notable amount of salt present the figures are too high. 3. Estimation of Sulphur in. Gunpozuder.-The method is analogous to that employed for dynamite, but pure benzene is used instead of alcohol. The saturation temperatures obtained in a series of determinations are represented by a curve, as in the case of nitro-glycerin. A curve is also plotted for the temperatures given by a mixture of alcohol and benzene with sulphur, the latter proceeding being adopted to prevent the crystallization of the sulphur. 4. Critical Temperatures of Industrial Oils.-The figures obtained with alcohol (specific gravity 0.8195) were as follows : Mineral oils (various), 135.5" to 140" ; valve oil, 197" ; animal oil, 120" ; sheeps' foot oil, 102" ; lard oil, 104" ; neats' foot oil, 95" ; colza, 132" to 135" ; Japanese fish oil, 108" C. As in the case of nitro-glycerin, the acidity of a fat lowers its critical tempera- ture. C. A. M.

ISSN:0003-2654    

DOI:10.1039/AN8962100237

出版商:RSC    

年代:1896

数据来源: RSC

摘要:

THE ANALYST. 243 INORGANIC ANALYSIS. Examination of Hydrofluoric Acid. I(. F. Stahl. (Zeit. angew. Chem., 1896, 225-229).-The impurities always present in the commercial acid are hydrofluosilicic acid, derived from the silica in the fluor spar, and sulphuric acid. These impurities influence the specific gravity of the hydrofluoric acid considerably, so that this con- stant does not alone give reliable information as to the purity of a sample of unknown origin. The method of analysis which the author has found most satisfactory is as follows : Thn, sample contained in a lead cylinder is placed in water at 50" C., and the specific gravity determined by means of a platinum hydrometer ; then, with a small platinum tube to serve as a pipette, three separate portions are weighed out : 1.Two grammes into a very small platinum crucible (holding about 5 c.c.). 3. Four granimes into a small platinum dish. A. TotaZAcidity.-The small crucible covered with a lid is placed in a platinum dish, into which is run from 25 to 50 C.C. of normal caustic soda according to the expected percentage of acid, the crucible is overturned, phenol-phthalein added, the acid and alkali mixed, and more alkali added, until after heating to about 50" C. a permanent red colour remains. B. HydroJEuosiZicic Acid.-To the acid in the large crucible about 5 C.C. of water are added, then slowly about 2 grammes of potassium carbonate, followed by 15 C.C. of 50 per cent. alcohol, and as many C.C. of 95 per cent. alcohol as of water used, which will bring the whole to a volume of about 25 C.C.containing about 50 per cent. of alcohol. After standing for an hour the gelatinous precipitate of potassium silico- fluoride is filtered off and well washed with 50 per cent. alcohol. I t is then placed, together with the filter paper, in a platinum dish with about 25 C.C. of water, warmed 2. Y 9 J 7 large , I ,, (holding about 40 c.c.). The number of C.C. used = a.244 THE ANALYST. to 50" C., and titrated with normal soda as in the determination of the total acidity. The number of C.C. used- b. C. Szdphzuic Acid--The acid in the platinum dish is evaporated under a hood until no more acid fumes are expelled, and the syrupy liquid remaining is titrated with normal soda solution, either phenol-phthalein or litmus being used as indicator.The number of C.C. used = c. The number of C.C. ( = a ) used in the first titration represents the alkali required to neutralize the hydrofluoric acid, hydrofluosilicic acid, and sulphuric acid. The number of C.C. used for the hydrofluoric acid alone = a , less the number used for hydrofluosilicic acid and sulphuric acid. But since in the potassium silicofluo- ride two atonis of fluoride were already neutralized, it would only have required b + And as 4 gramines of the substance were used for the estiniation of sulphuric acid, the number of C.C. there used must be divided Therefore the number of C.C. used for the free hydrofluoric acid is CL - ( i b + i), and since each C.C. normal alkali -- 0.020 gramme of HF, and two grammes of the sub- stance were used, this formula also expresses the percentage of free hydrofluoric acid.The percentage of hydrofluosilicic acid is giveu by the formula b x 1.8, and that of free sulphuric acid by c x 1.2. Other acids, like nitric or hydrochloric acid, are not likely to occur in commercial hydrofluoric acid, and if 53, would be readily detected. The following results were obtained with samples of hydrofluoric acid from various iiiakers. Specific gravity 1.299 1.264 1-253 1.244 1.864 1.282 1.247 1.234 Hvdrofluoric acid 39-6 42.2 44.3 48.1 48.6 51.1 54.2 48.6 33.5 b '2 C.C. to neutralize the free acid. by 3. The amounts of the constituents are in percentages. 1. 3. 3. 4. 5. 6. 7. 8. 9. Hidrofluosilicic acid 9.7 14.9 10.1 4.7 5.0 6.8 Free sulphuric acid 0.8 0.8 4.0 1-9 1.4 8.1 6.3 10.6 1.2 0.8 1.6 C.A. M. On the Precipitation of Calcium Oxalate. A. Lemoine. 1896, x . , 124-136.)-To obviate the difficulty of filtering calcium (BlLll. dc 1'-iss. belgc, oxalate precipitates which is often experienced, even when the precipitation has been made in boiling solution, the author describes the following method : A sufficient quantity of oxalic acid (preferably solid) is added to the calcium solution, followed by hydrochloric acid until the liquid becomes clear. The solution is boiled and then neutralized with ammonia, avoiding a large excess. The boiling is continued for about ten minutes, or until the greater part of the ammonia has been expelled. The precipitate obtained under these conditions is flocculent, subsides rapidly, and is easily filtered. I t may be washed with hot or cold water. It is claimed that in the majority of cases, even when the amount of calcium exceeds 0.5 gi-amme,~less than fifteen minutes is required for precipitation, filtration, and washing, whilst the comparative experiments described in the paper show that it is quite as accurate as the ordinarymethod.C. A. &I.THE ANALYST. 245 Resazurine as an Indicator. L. Crismer. (BUZZ. de Z’Ass. helge, 1896, x., 22-25.) -This substance gives an intense blue colour with alkalies, alkaline carbonates, and ammonia, whilst acids redden it. I t is more sensitive than turmeric, its solu- tions are stable, and it can be used with borax solution. A decinornial solution of borax, when titrated with the use of this indicator, exactly neutralized a decinormal solution of oxalic acid.On the other hand, it cannot be used for the titration of monobasic organic acids, is less sensitive to carbonic acid than phenol-phthalein, and is uncertain with nitric acid. Treated in alkaline solution with reducing agents (glucose, etc.), it gives a rose colour, with a vermilion-coloured fluorescence. It is prepared by dissolving 10 grammes of resorcin in 500 C.C. of ether, cooling to 50 C., and ,adding 8 grammes of fuming nitric acid diluted with ether. After several days, the crystals are collected, washed, and dried. For use as an indicator, 0.20 gramnie is dissolved in 40 C.C. of decinormal ammonia and the solution made Lip to one litre with water. C. A. IN. A Method for the Standardization of Potassium Permanganate and Sul- phuric Acid.(Ayncr. Clzem. JOZLY., 1896, xviii., 236-238.)-To obtain the potassium permanganate of normal composition the solution is prepared from the recrystallized salt, and the suspended manganese oxides removed by filtration. The solution is standardized by adding to a measured quantity of standard sulphuric acid a little hydrogen peroxide, and running in the permanganate so long as the colour disappears. More hydrogen peroxide is then added, and the titration continued to the same point. This is repeated until about 50 C.C. of the permanganate have been reduced, leaving in the solution a minute excess of hydrogen peroxide. Finally, the acid remaining free is titrated with decinornial ammonia, litmus being used as indicator. The quantity of permanganate in 50 C.C.of the solution is calculated from the formula : H. N. Morse and A. D. Chambers. 2KMn0, + 5H,O, + 3K,SO, = K,SO, + 2MnS0, + 8H,O + 50,. I n the test experiments described by the authors, 50 C.C. contained 0.16% gramme, whilst the amounts found by titration with potassium tetroxalate and oxalic acid were 0.1626 and 0.1628 gramme respectively. Sulphuric acid is standardized in the same manner with permanganate solution of known strength. In test experi- ments, using a solution containing in 50 C.C. 0.17435 gramme of potassium perman- ganate, the results were identical with those obtained by precipitation as barium sulphate. The hydrogen peroxide is the ordinary commercial solution, first rendered neutral by agitation with zinc oxide which has been heated in a muffle and then filtered through asbestos.C. A. M. Estimation of Potash. Ch. Fabre. (Comptes rendus, cxxii. [23], 1331.)-The author proposes to expedite the estimation of potash by the platino-chloride method in the following manner : The potash solution is evaporated on the water-bath with a slight excess of platinic chloride, any ammonia compounds being destroyed by a, few drops of nitro-hydrochloric acid. The residue is taken up by a little water and re-evaporated,246 THE ANALYST. then triturated with a little 90" alcohol, collected on a filter, and washed with alcohol till the washings are colorless, whereupon the alcohol in the residue is displaced by ether, which is itself driven off by evaporation, and the residue taken up by boiling water.The next step is to throw down the metallic platinum by means of mag- nesium, powdered and mixed with alcohol and water, and added gradually, in order that the reaction may not become too violent. 60" is the most favourable tempera- ture, and the formation of oxychloride of magnesium is prevented by adding a few drops of sulphuric acid at the end of the operation. After filtration, a slight excess of precipitated calcium carbonate, followed by potassium chromate, is added to the liquid, and titration is performed by decinormal nitrate of silver. The method is found exact to within 0.2 per cent. in the case of 48 per cent. of potash. c. s. Cuprous oxide as a Reagent for Nitrites. P. Sabatier. (Conzptes T C ~ ~ ~ Z L S . , cxxii. [24], 1417.)-The addition of a little red cuprous oxide to a dilute solution of sodium nitrite in concentrated sulphuric acid produces a fugitive, but intense, violet purple coloration, consisting of the cupric salt of nitrodisulphonic acid, and the same effect is produced (more gradually) by all the cuprous or cupro-cupric com- pounds, but not by cupric salts.This reaction is employed by the author as a test for nitrous acid, a drop of sulphuric acid being added to a drop of the suspected liquid, a few grains of cuprous oxide being then added to the mixture ; it is less sensi- tive than Griess' reagent, and is hindered by the presence of nitric acid in quantity. c. s. Natural and Industrial Phosphates-I11 : The Insoluble Residue. H. Lasne, ( A m . de CJzim. A7zuZyt., i. [ll], 207.)-The insoluble residue froin the hydrochloric solution (ANALYST, xxi., p.195) is calcined on a tared filter, the loss in weight indicating organic matter insoluble in acid. The calcined mass is usually white, containing only minute traces of iron, and is generally composed of granular and amorphous silica, with silicates of potassium, sodium and alumina. When fluorine is present in the phosphate, allowance must be made--at the rate of 0.3346 of silica to 1 of fluorine-for the silica volatilized as fluoride, provided there is an excess of silica present. I n other cases a weighed amount of silica (some 5 per cent.) should be added to the phosphate at the outset and deducted from the final results. The residue, being generally free from lime and iron, may be treat'ed by the usual methods for dissociating silicates, the simplest being to drive off the silica by hydro- fluoric acid, and take the loss in weight after the addition of a few drops of sulphuric acid and re-calcination.The new residue is dissolved in sulphuric acid and tested for alumina, alkalis, phosphoric acid (from insoluble amblygonite, monagite, etc.), titanic acid, and zirconium-the presence of rare elements affording assistance in determining the origin of the phosphate. Barium should also be looked for, its presence indicating adulteration of the (bone) superphosphate by plaster of Paris, which substance, according to the author's statement, contains a considerable quantity of barium sulphate. c. s.THE ANALYST. Volumetric determination of Mercury by Potassium Iodide.G. Deniges. (Bzill. SOC. de Pharm. de Bordenzix, April, 1896, p. 97 ; through Ann. de Cliinz. dnnlyt., i. [ll], 211.)-This method is based on the double deconiposition of mercuric salts and potassium iodide. A deci-normal solution of mercuric chloride is prepared by moistening 13.55 grainmes with 5 C.C. of hydrochloric acid, agitating with 10 C.C. of water and making the volume up to 1 litre. The iodide solution is made by dissolving rather more of the salt than is theoretically necessary, and titrating it with the mercuric solution in presence of a couple of drops of sodium bisulphite (36-40" Be). The addition of 1 C.C. of ammonia increases the stability of the reagent. Taking a as representing the number of C.C. of deci-normal mercuric solution required to titrate 10 C.C.of the iodide solution, 100 (a - 10) C.C. of water must be added to a litre of the iodine solution, in order that it may correspond to the &-normal solution of mercury. The corrections for dilutions of the reagent between :a and &,-normal strength are calculated as follows, a having the value already stated : between $= and ;,l,-normal ... 1 1 .. I ! 2 0 7 9 IjU 9 9 . . . (a -- 1 O ) y 1,000 x 4. X = 1 1 Y , CiJ 1 , ] 0 0 I , ... - -_ The results will correspond very closely to the theoretical quantities, provided the liquid is neutral or contains acid not exceeding 0.05 C.C. of HC1. The mercury is brought to the condition of chloride by the action of nitro- hydrochloric acid, there being practically no loss of mercury by evaporation, if the liquid is strongly acid and the operation is performed in a tube.To apply the test to a mercurial solution, 1 gramme of substance is placed in a test-tube with 2 C.C. of hydrochloric and 1 C.C. of nitric acid (double these quantities in the case of vermilion) and boiled until it has dissolved and z cloudy precipitate commences to form, whereupon 5 or 6 C.C. of water are added, and the whole made up to 100 C.C. The solution is titrated by noting the volume required to impart a feeble red coloration to 10 C.C. of the iodine solution containing two drops of bisulphite; if more than 14 C.C. are used, a correction, on the lines given above, must be made. By dividing the amount of mercuric solution into 1,000, the per- centage titre of mercury in the substance will be obtained.The results cited approximate very closely to the truth, the iodides of mercury, however, giving figures somewhat low on account of the formation of iodic acid, which is reduced by the bisulphite to an iodide of the alkali base, and falsifies the result of the titration. c. s. Decomposition of Alcoholic Indicator Solutions. Verbiese. ( J . fc( by. sucre, 1896, xxxvii., 20; through Chem. Zeit. Rep,, 1896, 171.)-The author remarks that acid solutions of phenolphthalein and other indicators in alcohol often become useless in a few days, and lead to erroneous results owing to the formation of acetic acid. F. H. L.248 THE ANALYST. Volumetric Process for the Determination of Phosphoric and .Arsenic (Nordisk plLawn. Tidsskrift, 1896, iii., 77 ; through Chem.Acids. A. Christensen. Xeit. Rep., 1896, 153 and 171.)-This process depends on the reaction : KBrO, + 6KI + 6H,,PO, = 6KH,PO, + 61 + KBr + 3H,O. For the practical analysis of phosphates 4 to 5 grammes of the sample are dissolved in 20 to 30 C.C. of 1.2 nitric acid, and diluted to 500 C.C. Twenty to fifty C.C. are treated with 1 to 2 grammes of silver nitrate, either solid or as a concentrated solution, and normal soda is dropped in until the yellow precipitate no longer disappears. Ten per cent. ammonia is then added till the liquid is alkaline, and the whole is boiled for a few minutes and allowed to settle. The precipitate is filtered off, washed free from lime (using at the last a 2 or 3 per cent. solution of potassium nitrate), and returned to the original flask by the aid of the same liquid. A solution of 0.3 to 0-6 gramme of common salt is introduced, and the mixture warmed slightly. The solution of the sodium phosphate is filtered into a stoppered flask, the residue washed till neutral with the nitrate liquid, 10 C.C. of seminormal sulphuric acid, 3 grammes of potassium iodide, and 10 C.C. of 5 per cent. potassium bromate added, and the whole made up to 100 or, at most., 120 C.C. After standing for half an hour at 40" or 50" C., it is ready for titration. In the analysis of basic slags the magnesium precipitate is washed first with 2.5 per cent. ammonia, then with 90 per cent. alcohol, rinsed into a stoppered flask, and 30 C.C. of decinormal sulphuric acid added for every 0.1 gramme of phosphoric anhydride it contains ; the liberated acid being estimated as above. In the case of arsenic acid, good results may be obtained by precipitation with magnesia mixture, Puller's correction for the solubility of the double salt in weak ammonia being employed. The volume of the liquid before titration must always be kept uniform. F. H. L.

ISSN:0003-2654    

DOI:10.1039/AN8962100243

出版商:RSC    

年代:1896

数据来源: RSC

摘要:

248 THE ANALYST. APPARATUS. Bunge’s Improved Analytical Balance. (Chenz. Zed., 1896, xx., 32G.)-In the ordinary process of weighing much time is wasted, and considerable harm is done to the agate bearings by the oscillations of the beam during the earlier attempts at attaining a balance. This improvement consists in the addition of a separate arrangement under the left-hand (object) pan, by means of which the weight of the substance is read directly to the nearest gramme on a scale attached to any convenient spot in the case. This pan is carried by a gilt steel stirrup made so as to expand in length according to the weight of the object. Underneath it is an agate point fastened to the arresting gear of the pan, which depresses to a greater or less extent an adjustable agate cup fixed to the shorter end of the lever-balance, the opposite extremity of which carries the pointer travelling over the scale.In the case of balances that are always swinging, enough space is left to permit of this inovement ; but in the other variety the addition is thrown in and out of gear by an eccentric. The improvement is patented in Germany. F. H. 11.THE ANALYST. 249 Araeometer Pipette. Greiner and Friederiche. (Zeit. anal. Chem. 1896, xxxv., 169.)-This consists of a peculiarly shaped pipette in the body of which is inserted a, small hydrometer of toughened glass. I t is especially suitable for the rapid determi- nation of the specific gravity of liquids, of which only a small quantity is availabie. Owing to the small size of the hydrometer the scale is necessarily limited.C. A. M. Standardizing Volumetric Apparatus. L. L. de Koninck. (Chenz. Zed., xx., 1896, 460.)-The author draws attention to the inexact way in which measuring cylinders are usually graduated, and insists on each piece of apparatus being tested before use. He observes that burettes, instead of being marked in true cubic centi- metres, are often divided into what he calls milliniohrs-the apparent volume of 1 gramme of water at 15" C., as weighed in the air with brass weights (100 milli- inohrs = 100.2 c.c.). For a quick test as to whether due care has been taken in the graduation, he suggests that the distance between the 0 and the 10 mark should be measured with a pair of compasses. If the other corresponding intervals are the same, the tube may at once be rejected, for it is impossible to manufacture glass tube with strictly parallel sides.For the calibration of tubes closed at one extremity, the process of inverting them and allowing water to run in from a standardized burette is useless, since the meniscus, being inverted, causes the volume of the liquid introduced to appear quite different from its true amount. To perform the operation correctly, the cylinder should be provided with a rubber cork having two holes. Through one is inserted a tube of 1 m.m. internal diameter drawn out to a fine point inside the vessel, bent upward again to a U-shape on the outside, and closed with a rubber tube and screw- clamp. The second hole carries a straight tube of the same size, drawn out to a point after the manner of the jet of a, Mohr's burette, The vessel is completelyfilled with water, and the stopper carefully inserted so as not to introduce any air. It is next fixed in the normal position, and the clamp opened until the former tube is exactly filled with air. Successive volumes of liquid may then be run off' either into a tared beaker and weighed, or simply into a correctly graduated burette. The process can be modified in a variety of mays, and if proper care is exercised in keeping the temperature normal, it will be found very exact. F. H. L.

ISSN:0003-2654    

DOI:10.1039/AN8962100248

出版商:RSC    

年代:1896

数据来源: RSC

摘要:

THE ANALYST. 249 R E V I E W . ‘‘ PHINCIPLES AXD PRACTICE OF AGRICULTURAL AXALYSIS.” By HAHYEI- W. WILET. Volume I. : Soils. Volume 11. : Fertilisers. Chemical Publishing Company, Easton, Pa. IN March, 1894, we had the pleasure of noticing the first monthly number of a new work by Prof. Wiley, bearing the above title. The treatise, as might have been expected, has outgrown the proportions anticipatorily assigned to it by the author, and is not yet complete. We have before us, however, the two volumes already2 5.0 THE ANALYST. completed, tht! first dealing with soils and the second with fertilisers. The chemistry and analysis of farm products and feeding stuffs is deferred till the third volume, still in course of publication. The two volumes already published do not disappoint the promise given by tlie work in its infancy.The volume on soils is a storehouse of valuable information, not the least acceptable part of which is to be found in the liberal bibliographical refer- ences appended to each part, each reference being connected by a key number with the portion of the text to which it relates. -4s an instance of the conscientious spirit in which the book is written, it may be mentioned that no fewer than twenty pages are devoted to the various methods of collecting samples of soil, and to the various considerations affecting this important initial step towards soil analysis Such frequently neglected subjects as soil thermo- metry are also treated. Methods for investigating the cohesive and adhesive properties of soils, their porosity and absorptive powers, and their capillary properties as regu- lating water movement, as well as their permeability, both as regards water and air, are discussed at length.So also are modes of investigating the selective absorptive potver of soils for potash and other chemical constituents of manures. The subject of mechanical analysis also occupies a great deal of the author’s attention, and we have here, for the first time, brought together a summary of modern improvements upon the old crude methods of mechanical analysis directed towards obtaining results possessing the advantage of strict comparability as well as of practical meaning. A Very beautiful set of micro-photographs of sand of various qualities and degrees of coarseness, silt, clayey particles, etc., appropriately illustrates this part.Of the nearly six hundred pages comprising the volume on soils, the first three hundred are almost entirely devoted to such physical aspects of soil examination as have been referred to, the remaining half being devoted to chemical analysis. The portion on the chemical analysis of soils deals not only with the determination of the total percentages of the various constituents, but also with the methods devised by various investigators for the attempted estimation of the L ‘ available ” constituents, such as the quantities soluble in water, carbonic acid water, weak ammonia salts, acetic acid, weak citric acid, etc. The determination of the nitrifying power of soils, as based upon bacteriological culture methods, and the investigation of the distribution of the nitrifying organisms in the various layers of the soil, are discussed at some length in a special part, along with the origin and estimation of oxidised nitrogen in soils, and in rain and drainage waters. Any agricultural chemist who can read through the volunie on soils without adding to his stock of knowledge, must be enviably well versed in his subject. I n the volume on the analyses of fertilisers, we have brought together, in a compact space, a large number of processes which have obtained inore or less pro- fessional recognition.The actual practitioner in agricultural analysis will be glad to see collated a considerable number of the arbitrary ammonium citrate ” methods for the determination of reverted and available phosphates in various countries, as laid down by custom or by oficial convention.Prof. Wiley proposes that the words “ available ” and “ reverted ’’ should be abolished, as sometimes begging an unsettled The gases in soils also come in for much detailed attention.THE ANALYST. 251 question, and suggests the use of the words ‘‘ citrate soluble ” to describe the phos- phates dissolved in such processes. I t is only necessary, however, to glance through the records which he has brought together of experiments with such methods, to realize how variable are the results yielded, depending as they do, not merely upon the acidity or alkalinity of a difficultly neutralizable reagent, but also upon the actual quantity of reagent used, and upon the time and temperature of digestion.He very properly points out that comparative results are only to be obtained by rigid adherence to the details of the particular process of which circuiiistances may dictate the choice. Prof. Wiley appears to omit reference to what some chemists venture to consider a much more satisfactory reagent than ammonium citrate, viz., a weak (I per cest.) solution of citric acid, as suggested by Tollens and Stutzer a dozen years ago. Nore recent investigation has shown this solution to have approximately the acidity of the root sap of plants. Its application to soils is, as has already been mentioned, dealt with in the volume on soils, but its applicability to the analysis of fertilisers has either been overlooked by the author, or-as seems inore probable-he considers that the aniiiioniuin citrate process, modified in some way or other, is too firnily establi~hed for displacement by the more simple and apparently more rational citric acid method.I n a short section on the manufacture of superphosphate, it is satisfactory to see that the author does not repeat the old text-book error, still fashionable, of stating that the soluble phosphate ” in superphosphate consists merely of tetra-hydric monocalcic phosphate, but points out that, as first shown by Ruffle, when excess of acid is used, as in ordinary manufacture, free phosphoric acid constitutes at least a considerable part of the soluble phosphatic matter. The matter on nitrogen determination is necessarily and properly voluininous, and includes descriptions of the various modifications of Kjeldahl’s process.Perhaps the least satisfactory feature of the book is the section on the deter- mination of iron and aluminium oxides in raw phosphates. A number of methods are given, all of which, excepting that of Hess, direct that the raw material shall be dissolved in nitric acid, with or without the addition of hydrochloric acid. This serious error, which is involved in the Glaser as well as other processes, results in the inclusion of iron pyrites with oxide of iron, a substance to which it is not techno- logically analogous, inasmuch as pyrites is scarcely, if at all, attacked by the acids used in superphosphate making, whereas iron oxide is so attacked, and vitiates the keeping power of the superphosphate.This was fully pointed out by Shepherd in THE ANALYST for November, 1893, and English analysts cannot fail to regret that Prof. Wiley has not drawn attention to this very important matter, particularly as neglect of this consideration has led to serious under-valuation, more particularly of certain descriptions of Anierican phosphates, which contain considerable quantities of perfectly harmless pyrites. As such phosphates form a subject of international trade between England and America, it seems a pity that, in a work which will take so highly authoritative a place as that by Prof. Wiley, this source of error should not have been pointed out, with a view to inducing American analysts to assimilate their practice to that of English chemists by using simple hydrochloric acid as a solvent, and thus to difl’erentiate between pyrites and iron oxide.I t is also to be regretted that the only modification of the acetate method for252 THE ANALYST. oxide of iron and alumina, is the somewhat cumbrous and not very satisfactory one of Hess. The acetate method has been largely discarded, not on account of any inherent untrustworthiness, but on account of the loose and unwarranted assumption, so long made with regard to it by many chemists, that the ammonium acetate pre- cipitate consisted merely of ortho-phosphates of iron and aluminium. Experienced employers of the process, who have not fallen into the error of ignoring the presence of lime in the ammonium acetate precipitate, and who, after weighing every precipitate, have taken the precaution to actually determine the lime and phosphoric acid in it, have never had to complain of its untrustworthiness, though no doubt the accurate analysis of the precipitate necessary for correctly arriving at the results involves somewhat nicer manipulation, and is somewhat more tedious than the Glaser process.I t almost invariably gives results identical with the Glaser process-provided, of course, that the Glaser process is modified by the substitution of hydrochloric for nitric acid or nitro-hydrochloric. I t may further be remarked that it appears to be not satisfactory, in these days of accurate commercial analyses, that chemists should allow themselves (except for rough and ready work in the factory) to assume that half the weight of the mixed phosphates of iron and aluminium consists of the bases in question.Aluminium phosphate contains 42 per cent. of aluminium oxide, while the percentage of ferric oxide in ferric phosphate is nearly 53. I n commercial contracts, the permissible limits of iron and aluminiuin oxides in phosphates are rigorously laid down, and even a few tenths per cent. materially affect the invoice value of the cargo of phosphate. The assumption that exactly half of the mixed “Glaser” phosphates consists of iron and aluminium oxides-altogether irrespectively of the proportions in which these two bases are present-may well be sufficient to lead to consequences financially serious to either buyer or seller, and even to the rejection of unfair cargoes. The right of rejection often lying with the buyer if a given limit, such as 4 per cent.of the mixed oxides, is overstepped. I t does not involve much time or trouble to determine the phosphoric acid in the Glaser precipitate, and thus obtain correctly the real quantity of the oxides with which it is combined. I t is, however, not given, even to so careful and experienced an author as Prof. Wiley, to include in one treatise all that may be desired on every subject of which he treats, and attention is drawn to these points in the hope of seeing thein dealt with in a future edition, and not, in any spirit of inappreciation of a work which can scarcely fail to take an indispensable place amongst the standard books of reference in the library of the specialist. As a practical guide-book, it is perhaps scarcely a book for the beginner ; for while the author details processes, and states when and where they have been officially recognised, he leaves the operator to make his own selection, somewhat after the manner of Fresenius. To presuppose that his choice will be wise, is to presuppose the existence of experience and the faculty of acquired practical intuition that experience alone develops. Analysts, both in England and in America, owe Prof. Wiley their gratitude for these two volumes, and will look forward with pleasurable anticipation to the com- pletion of the third volume. B. n.

ISSN:0003-2654    

DOI:10.1039/AN8962100249

出版商:RSC    

年代:1896

数据来源: RSC