摘要:

JUNE, 1913. VOI. XXXVIII., NO. 447. THE ANALYST. PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS. THE monthly ordinary meeting of the Society was held on Wednesday evening, May 7, in the Chemical Society’s Rooms, Burlington House. The President, Mr. L. Archbutt, F.I.C., occupied the chair. The minutes of the previous ordinary meeting were read and confirmed. A certificate of proposal for election to membership in favour of Mr.F. S. Sinnatt was read for the second time ; and certificates in favour of Messrs. Charles William Birch, 550, Simcoe Street, Victoria, B.C., Public Analyst for the City of Victoria, British Columbia ; Ulick Richardson Evans, B.A. (Cantab.), 28, Victoria Street, Westminster, Consulting Electro-Chemist ; and Harold Warren Gill, B.Sc.(Lond.), A.I.C., South African School of Mines, Johannesburg, were read for the first time. Messrs. H. F. E. Hulton, A.I.C., C. T. Kingzett, F.I.C., and A. J. Parker, were elected members of the Society. The following papers were read : 4 4 A New Apparatus for Maintaining Constant Temperatures,” by F. H. Dupr6 and P. V. Duprb, F.I.C. ; c 4 On the Proportionate Determination of Cocoanut Oil and Palm-Kernel Oil in Mixtures,” by H.R. Burnett and Cecil Revis ; (( Examination of the Oils from Manihot ceara and Funtumia elas- tics, and a Comparison of their Properties with those of Linseed and Hevea Oils,” by S. Rideal, D.Sc., F.I.C., and L. H. D. Acland, B.A. ; 4 6 The Recovery of Iodine from Residues,” by Harold W. Gill, B.Sc. ; and (‘ The Composition of Milk,” by H. Droop Richmond, F.I.C. OBITUARY. WE greatly regret to record the death, on April 29, of Dr. M. A. Adams, of Maid- stone, a past President of the Society. An obituary notice will appear in the next issue of the Journal.

ISSN:0003-2654    

DOI:10.1039/AN9133800241

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

242 TRIMEN: EGYPTIAN BUTTER AND SAMNA EGYPTIAN BUTTER AND SAMNA. BY S. H. TRIMEN. (Read at the Meeting, April 2, 1913.) IN Egypt, as in all Eastern countries, butter-fat is used by the native population, and not butter as it is known in Europe, This butter-fat, called “ghee” in India, is known in Egypt under its Arabic name ‘‘ samna ” (or more correctly samn ”). In the course of the following paper the commercial product will be referred to under that name.Samna is usually made in the following way: The whole milk is churned by shaking it up in goatskins until the butter (Arabic, “zibda”) forms. This part of the process is done by the fellahin in their villages ; but because it is not economical to make samna in small quantities, they do not make it themselves, but sell the zibda to the samna makers.These latter usually live in the towns, and send their agents round the neighbouring villages to buy the zibda from the fellahin. The zibda which has been collected in this way is heated in large pans until it melts; the water, casein, etc., subside, the samna is poured off and usually stored in old kerosene tins. The chief source of Egyptian samna is the milk of the buffalo (“ gamoosa”).Cow’s milk is probably used only to a limited extent, the number of cows in Egypt at the present time being comparatively small, and their milk being largely used for drinking purposes. Although there are large numbers of sheep and goats in Egypt, their milk is used almost exclusively for making cheese, and is not used to any extent in the manufacture of samna.It is quite likely that from time to time small quanti- ties of zibda from cow’s, sheep’s and goat’s milk go to the samna makers, but the amount is apparently not large enough appreciably to affect the analytical constants of normal Egyptian samna, although it may account for an occasional abnormal sample. Syrian samna, which is largely imported into Egypt (1,073,694 kgrms.in 1910 and 676,051 kgrms. in 1911), is, on the other hand, generally believed to be made from sheep’s and goat’s milk. I t s actual composition will be discussed later, but it is unlikely that cow or buffalo milk is used in any quantity in its manu- facture. Indian ghee, which is also imported into Egypt, is apparently made from buffalo’s milk, though cow’s milk may enter occasionally into its composition.Samna is only ‘made on a large scale during the winter months, and it is the custom amongst the more wealthy of the Egyptians to purchase their annual supply at that season and store it in a cool place during the year. Being for all practical purposes a pure fat, its keeping properties are, of course, vastly superior to those of butter ; indeed, it is not uncommon for samna to be used as a food some years after its manufacture, but reference to this will be made later on.Samna as it is known in Egypt is pale yellow in colour, possessing a smell, sometimes cheesy, sometimes acid, frequently rancid, and always distinctly un- pleasant to European ideas. Indian ghee is very similar, but Syrian samna frequently has a not unpleasant smell suggestive of ethyl butyrate ; the writer, however, hasnot been able to discover whether this smell is natural to the samna or due to the addi-TRIMEN: EGYPTIAN BUTTER AND SAMNA 243 tion of a small quantity of an ester.During the cold weather samna is a solid opaque greaae with a granular consistency ; in summer it is sometimes liquid, but is usually in a semi-liquid state.Butter proper, as distinct from samna, is made in Egypt all the year round, but only in the large towns, for the use of European resi- dents. As buffalo butter is very pale in colour, it is the usual practice for the dairy- man to make it resemble cow butter by the addition of a small quantity of butter colouring. Apart from such sophistication which may, or may not, be justifiable, the adulteration of butter and samna is common in Egypt.Cocoanut oil and clarified animal fat are the usual adulterants, and enormous quantities are imported into Egypt for this purpose--1,397,139 kgrms. of cocoanut oil were imported in 1911-in fact, a large recognised trade exists between the importer and the samna and butter maker, fostered by the former’s assiduous and plausible agents.Both sesame and cottonseed oil have been detected occasionally in adulterated samna; it is not known, however, whether they are added directly to samna, or whether they enter it as ingredients of imported butter adulterants, possibly ear- marked ” by their means before exportation. A few examples of adulterated samna are given in Table I., and these samples, which are typical, show the large proportion of adulterant usually added by the Egyptian sophisticator.Since the examination of butter and samna in Egypt presents many features not encountered in normal analytical practice, it was not at first easy to decide which of the constants it would be most advantageous to determine in routine work. At the present time the writer determines the Reichert-Meissl, Polenske, saponification and Valenta values, the Zeiss butyro-refractometer reading at 40’ C., and tests qualitatively for sesame and cottonseed oil.As the ideal method of testing butter is still to be found, this list is in no way final. Other and newer methods have been tried in the hope of finding a routine test more particularly suitable to the local conditions, but without much success.An account of the trial of these methods with samna is now given : The Av8 Lallement (Baryta) Method.-This method was tried on some forty samples of samna, butter, possible adulterants, and mixtures containing adulterants, Some of the results, together with the other analytical constants determined, will be found in Table I.I n every case genuine Egyptian butter and samna were found to give negative values for b - (200 + c), thus showing that Lallement’s statements apply to butter-fat from the buffalo as well as that from the cow. Unfortunately, however, these values are in many cases so low that a negative value is still given after the addition of a considerable amount of cocoanut oil. It has been found that the addition of 10 per cent.cocoanut oil to a samna increases the value of b-(200+c) on an average by 6.2. This means that if the b-(200+c) value of the samna is less than -6.2, the samna will still give a negative figure for this value after the addition of 10 per cent. cocoanut oil, and it is obvious that when the b-(200+c) value is as low as - 20, as it sometimes is, as much as 30 per cent.cocoanut oil could be added with- out its detection by this method. It is also stated by Revis and Bolton (Allen’s *‘ Commercial Organic Analysis,” fourth edition, vol. ii., p. 289) that in cases where negative values are obtained for b-(200+c) with mixtures of cocoanut oil andTABLE I. NO. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16" 17" 18" 19" 20" 21* 22 23 24 25 26 27 28 29 30 Description. ~ Genuine Egyptian ,mmm .. . ) I 8 , ,) ... 9 9 9 , ) I ... 9 , 9 , 2, ... 9, I ? 9 , ... 9 9 9 , 9 , ... 9 9 Y, > 7 ... No. 7 + 10 per cent. cocoanut oil No. 7 +20 per cent. ,, No. 7+30 per cent. ,, Genuine Egyptian samna . . . No. 11 + 10 per cent. cocoanut oil Genuine Syrian samna . . . ... I I P 9 9 ) * * * ... Goat butter-fat ... ...... 3, 3 , ... ... ... Sheep butter-fat . . . ... ... Buk,lo bdler-fat' ... ... ... ... Butter-fat from milk of one cow ADULTERATED SAMNA. Bought in Cairo ... ... Bought in Alexandria . . . 9 9 9 9 32 9 ) ... I . . ... ... ... BUTTER ADULTERANTS. Cocoanut oil ... ... ... Mutton-fat . . . ... ... Lard ... ... ... ... Cottonseed oil ... ... ... Sesame oil ... ... ... 42 r-4 2 % 0 '8 $4 31.9 34.1 33.4 33.3 33.0 32.1 31.6 29-6 27.4 24.9 36-6 33.7 24.4 27.2 31.2 20.8 22.9 22.9 26.5 32.3 23.4 10.4 18.2 12.8 5.0 7.2 - - - - 2.2 1.5 1.7 2.1 2.0 1.4 1.9 2.8 4.1 5.0 2.3 3.1 4.1 5.3 6-4 6.5 4.9 2.8 3.1 1.6 1.6 8.5 7.3 10.3 4.3 16.9 0.3 0.3 - - 41.7 42.1 42.2 42.0 41.7 42.0 42.0 41.3 40.6 39.8 41.7 40.6 42.9 43.0 42.3 40.7 43.6 44.0 44.0 42.0 46.5 39.9 40.0 39.0 43.7 35.0 47.2 50.0 58.8 61.6 231.3 230.0 230-9 229.3 228.3 228.1 230.9 232.6 235.6 237.3 236.3 237.4 227.3 232.8 235.5 231.6 224.5 223.2 216.6 228.4 217.1 233.5 236.3 238.5 216.6 254.7 194.5 193.1 185.3 193.8 a.316.2 314.5 315.6 313.5 312.1 311.8 315.8 317.9 322.1 324.4 323.0 324.6 310.7 318.2 322.0 316.6 306.9 305.1 296.1 312.2 296.8 319.2 323.0 326.0 296.1 348.2 265.8 264.0 253.3 264.9 Bary t a Value.b. 256.0 253.1 255-3 253.0 252.9 253.8 248.9 252.9 259.1 265-3 251.5 256.2 254.7 254.6 253-9 267.1 253.5 256.4 253.0 251.8 251.8 282.9 275.6 282.3 276.1 296.9 264-0 263-7 235.1 247.1 c. 60.2 61.4 60.3 60.5 59.2 58.0 65.4 65-0 63.0 59.1 71.5 68.4 56.0 63.6 68.1 49.5 53.4 48.7 43.1 60.4 45.0 36.3 47.4 43.8 20.0 51-3 1.8 0.3 18.2 17.8 b - (200 + C.) - 4.2 - 8.3 - 5.0 - 7.5 - 6.3 - 4.2 - 16.5 - 12-1 - 3.9 + 6.2 - 20.0 - 12.2 - 1.3 - 9.0 - 14.2 + 17.6 + 0.1 + 7.7 + 9.7 - 8.6 + 6.8 + 46.6 + 28.2 + 38.5 + 56.1 + 45.6 + 62.2 + 63.4 + 16.9 + 29.3 % SO 8 m .FIG * - - - - trace - trace + + - * Prepared in laboratory.TRIMEN: EGYPTIAN BUTTER AND SAMNA 245 butter-fat that the value of b for these mixtures always exceeds 260. From theoretical considerations alone, it is difficult to see how this can always be the case. I n genuine butter-fat b is frequently below 250, and according to AvB Lallement 10 per cent.cocoanut oil only increases the value of b by from 4.5 to 5. Hence if 10 per cent. cocoanut oil be added to a butter-fat, with, say, a b value of 250, the value of b can only be increased to a maximum of 255-i.e., considerably below 260.It is also curious to find on the very page on which this statement is made an example given of a butter containing 10 per cent. cocoanut oil with b value of 259.2. However these diecrepancies may be explained, the statement that mixtures of cocoanut oil and butter-fat always give values for b, exceeding 260, does not hold good for samna. I n Table I. will be found two samples containing 10 per cent.cocoanut oil (Nos. 8 and 12) and one containing 20 per cent. (No. 9), all of which give lower b values than 260. In his original paper AvB Lallement states that the method breaks down with overheated or rancid butter-fat. I t is possible, therefore, that the failure of the method in these cases may be due to the samna having been overheated in its manufacture. Whether this be so or not, it is difhult to understand how any butter-fat with a maximum figure of -23.8 for the b - (200 + c ) value, such as is given by Avd Lallement, can ever give a positive reading for that value after the addition of a small quantity of cocoanut oil.Another fact which detracts from the value of this method in an Eastern country is that both goat and sheep butter-fat give at times positive values for b - (200+c).The figures (Table I., Nos. 16 to 19) were obtained from butter made in the Iaborabory from mixed genuine milk of sheep and goats. Two samples of each were examined. The two samples of goat butter-fat gave +17.6 and +0*1 for the b - (200+ c ) value, and the two samples of sheep butter-fat + 7.7 and 9-7 respectively.I t is not claimed that these results are in any way typical; they are simply to show that the method does not always hold good with butter-fat other than that of the cow and buffalo. Curiously enough, the only samples of Syrian samna-believed to be made chiefly from sheep milk-examined by this method all gave negative values for b - (200 + c), viz., - 1.3, - 9-0, and - 14.2 (Table I., Nos.13 to 15). I n .view of the many shortcomings of this method, which have been sufficiently indicated above, the writer cannot endorse the high opinion of its value which has been expressed by Revis and Bolton (ANALYST, 1911, 36, 393). The writer is of the opinion that the information given by it is not sufficient to justify its adoption as a routine method for the testing of butter and samna in Egypt, but that as a con- firmatory test in suspicious cases it might occasionally be of value.The Shrewsbury-Knapp Method-In using this method the writer, in common with other workers (ANALYST, 1911, 36, 334, and 1912, 37, 6) experienced great difficulty in obtaining concordant results. This difficulty was doubtless in part due to the variable temperature experienced.Indeed, it seems improbable that a method which depends on the solution of a substance at a definite temperature can ever be employed as a routine method in a country such as Egypt, where it is no uncommon occurrence for a room temperature to vary 25" F. in the course of a morning, The modification suggested by Revis and Bolton (ANALYST, 1911, 36, 534) has not been tried up to the present,246 TRIMEN: EGYPTIAN BUTTER AND SAMNA The Fendler Alcohol S o l ~ b i l i t ~ Value (ANALYST, 1910, 35, 355).- The writer i not at present in a position to give a definite opinion on this method, although several determinations have been made.I n common with the Shremsbury-Knapp method, its chief defect is that it depends on obtaining a solution at a constant temperature, which naturally militates against its use in a hot climate.I t is hoped that the introduction of certain modifications to remove the temperature difficulty will render it a valuable confirmatory test. I t s great advantage is that, suitably modified, it is very little trouble to carry out, being merely a continuation of the Reichert method, no additional quantity of butter-fat having to be weighed out.Colour Test for Cocoanut Oil.-Ludwig and Haupt’s qualitative test (Zeitsch. Untersuch. Nahr. Genzissm., 1907, 13, 605-610) was tried, but unfortunately with negative results. The chief object in trying the above methods was to find one that would enable the analyst to distinguish readily between different varieties of butter-fat. To show how extremely useful it would be to have such a method, it is only necessary to point out that the analytical constants of sheep butter-fat (vide Syrian samna) resemble very closely those that would be given by a mixture of cow or buffalo butter-fat with cocoanut oil.So far, unfortunately, no quantitative method has been found by which different kinds of butter-fat can be detected.Very useful indications, however, have been given by Hinks’ test (ANALYST, 1907, 32, 160), a, qualitative method originally devised for the detection of cocoanut oil in butter. Different varieties of butter-fat and mixtures of butter-fat with cocoanut oil were examined by this method, and it was found that the final products, when viewed under the microscope, dieer with different kinds of butter-fat.The following is a description of the microscopical appearance of the various samples examined : Bufalo Butterfat (Egyptian Samna). - Small yellow globules, more or less rounded and homogeneous, but with no crystalline structure. Cocoanut Oil.-Fine white needles, growing concentrically, markedly resembling the conventional geographical map-sign for mountains. Bufalo Butterfat + 10 per cent.Cocoanut Oil.-Small yellow globules, most of which are surrounded with small hair-like crystals, which appear to be growing out of the globules. Their appearance resembles that of a ciliated unicellular organism. Bufalo Butter-fat + 50 p e r cent. Cocoanut Oil.- The appearance approximates very closely t o t h a t of pure cocoanut oil, the difference being that the crystals are yellowish not white; there is also a certain amount of amorphous material besides the crystals.Goat Butter-fat.--A very similar crystalline growth to that from cocoanut oil, only that the crystals are smaller and slightly yellowish. It strongly resembles a butter-fat containing a large percentage of cocoanut oil. Sheep Butter-fat.-Globules, larger, lighter in colour, and more homogeneous than is the case with buffalo butter-fat.They have no crystalline structure, and are frequently split across in a manner resembling some of the larger varieties of starch granules. Syrian Samna.-The two samples of Syrian butter-fat examined by this methodTRIMEN: EGYPTIAN BUTTER AND SAMNA 247 both gave results identical with that obtained from sheep-butter, no trace of crystalline structure being observed.These results appear to justify the following conclusions : 1. I n the absence of cocoanut oil, it is possible by this method to distinguish between buffalo or cow butter-fat, and that from sheep and goats. 2. Owing to the close resemblance that exists between the product from cocoa- nut oil and goat butter-fat, it is impossible to detect the presence of cocoanut oil in goat butter-fat or the presence of goat butter-fat in cocoanut oil.Either being absent, the test is of considerable value as a confirmatory test for the other. If this were always the case, and if goat’s milk were never a source of Syrian samna, then its adulteration by means of cocoanut oil could be confirmed by this method.Unfortunately, it is generally believed that goat as well as sheep butter-fat enters largely into the composition of Syrian samna, and until a large number of samples have been examined this cannot reasonably be disproved. The Hinks’ method, unfortunately, has one great disadvantage, which detracts considerably from its usefulness. Owing to the low melting-points of the globules and crystals, which have .to be examined microscopically, it is practically impossible to use the method in hot summer weather.A cooled stage was tried, but was found to be quite impracticable owing to the condensation water which immediately formed on the cover-glass, completely obscuring the object. The test is, therefore, only of value in Egypt during the winter months.Before considering the analytical constants of butter-fat in Egypt, it is as well to state that the Reichert-Meissl, Polenske, saponification, and Zeiss values, are all strictly comparable with those obtained in English analytical practice. The Valenta values were, on the other hand, obtained by using acetic acid of sp. gr. 1.060, which happened to be the strongest glacial acetic acid in the laboratory when the tests were started.It has been used for all the tests, so that the results, although com- parable among themselves, cannot be compared with results obtained by using acid of a different strength. I n Table 11. will be found the mean results obtained from Egyptian butters and butter-fats during the past few years, arranged according to their ascending Reichert- Meissl values.All the results here tabulated were obtained from samples believed to be genuine : The Reichert-NeissZ Value, it will be noticed, is much higher than that normally obtained from cow butter, and shows a variation from 28 to 38. These are limiting values, the normal value, which will be discussed later, lying within closer limits.The Polenske Value varies directly with the Reichert, the variation being, however, very much less than in the case of cow butter. The minimum value observed by the writer was 1.1, and the maximum 2.4 ; both theso values, however, are abnormal. A variation of 10 in the Reichert-Meissl value was found to correspond with an average variation in the Polenske of 0.7, as against 2.0 with cow butter.Such a comparatively small variation considerably enhances the value of the Polenske figure in the detection of adulteration. Taking the mean of a number of experiments, the writer has found that the addition of 10 per cent. cocoanut oil to a butter-fat 3. Some Syrian samna is apparently made entirely from sheep milk.248 TRIMEN: EGYPTIAN BUTTER AND SAMNA - 42.0 42.5 42.5 42.0 42.0 42.0 42.0 TABLE 11.Mean Results from Egyptian Samna and Butte?., arranged according to the Reichert-Meissl Values, - 64.5 67.0 62.0 58.5 57.0 59.0 60.0 Number of Samples. 3 5 7 9 14 13 14 5 6 3 Reichert- Meissl (Limits). 28-29 29-30 30-31 31-32 32-33 33-34 34-35 35-36 36-37 37-38 Yolenske. 1.3 1.6 1.5 1.8 1.8 1.7 1.8 2.0 2.0 2.0 Zeiss a t 40" C. 42.5 42.9 42.1 42.1 42.1 42.0 42.0 42.1 42.0 41.8 Valenta (Acetic Acid, Sp.Gr. 1.060). 70.5 66.0 67.0 62-0 61.0 61.0 59.0 58.0 53.0 53.0 Saponification Value. 225.0 225.9 227-8 229.5 229.5 229.8 230.9 231-6 233.8 234-9 raises the Polenske value by 1.4-that is, by twice the normal variation of the Polenske value. Hence the detection of small amounts of cocoanut oil in butter and samna of known Egyptian origin becomes a comparatively easy matter.The Sapon@cation Value varies directly and very regularly with the Reichert-Meissl figure. Plotted graphically, it gives what is practically a straight-line curve. Such a regular variation is, of course, very useful in forming an estimate of the genuine- ness of suspicious samples. The Valenta Value varies inversely and fairly regularly with the Reichert.As will be seen from the table, the variation is considerable, and were the analyst unaware of it he might frequently be unduly suspicious of abnormal samples. The Zeiss Butyro-Refractometer ; Except at the extreme limits, where slight variation is shown, this value is practically constant at 42. TABLE 111. Results showing Seasonal Variation in the Composition of Egyptian Butter.I I I August ... September October . . . November . . December . . . January ... February ... March ...' April ... May ... 3 2 5 4 5 4 4 4 4 - 30.5 30.9 33.2 32.0 35-0 35.2 34.4 34.0 33.6 - ' 1.3 1.3 1-4 1.6 1.9 1.9 1-9 1.8 1.7 I I I sp. gr. 1.060). 42.0 67.5 1 227.3 42.0 I 64-0 I 228.5 - 229.9 228.8 231.5 231.6 231.0 231.1 231.1TRIMEN: EGYPTIAN BUTTER AND SAMNA 249 Seasonal Variation.-In such a country as Egypt it is only to be expected that the composition of butter-fat will vary considerably during the course of the year.That this is the case is shown in Table 111. The figures there given are the mean results of weekly samples of butter (not samna) kindly supplied by Mr. Howie of the Cairo Hygienic Dairy. The information it supplies, however, although incomplete, is sufficient to indicate the general tendency of the variation which takes place during the course of the year.I t shows (1) that the Reichert-Meissl, Polenske, and saponification values are at their lowest in the summer, and at their highest in the winter ; (2) that the Valenta value varies inversely with these values-that is, it is highest in the summer and lowest in the winter ; (3) that the Zeiss reading remains practically constant.These changes are to some extent, no doubt, directly due to the effects produced on the animals by the great seasonal variation of temperature, but it seems probable that they are mainly brought about by differences in diet. During the summer months in Egypt there is an almost entire absence of green fodder, the cattle being mainly fed on chopped straw (“ tibn ”), while in winter there is practically an unlimited supply of green food. The chief green fodder of the country is a lucerne (Medicago sativa), known locally as ‘ I berseem,” which flourishes from November to May.I t is during these months that the cow-buffaloes give their maximum and richest supply of milk, and it is then that most of the year’s supply of samna is made.On looking through Table III., it will be seen that during these months the analytical constants show comparatively little variation, and it is therefore reasonable to suppose that normal Egyptian samna, when it is fresh, will show similar slight variations. This is undoubtedly the case, the normal variations being as follows : Reichert-Meissl, 30 to 36 ; Polenske, 1.4 to 2.0 ; saponification value, 228 to 232 ; Valenta, 57 to 64 ; and Zeiss practically constant at 42.The Alteration in Composition of Sarnna on Keepiy.-Crispo (Bull. Soc. Chim. Belg., 1910, 24, 436-438) has pointed out that on keeping butter-fat in closed flasks for a period of one to three years there is usually a slight gain in the proportion of volatile acids (Reichert-Meissl values).Before Crispo’s paper appeared, the writer had had occasion to examine certain samples of samna which gave unusually high saponification values and correspondingly low Valenta values, but which were other- wise normal, and apparently genuine. As the samples all had a rancid smell, it was thought that the abnormal figures might possibly be due to the effect of age.I t was therefore decided to keep a few samples of fresh samna in glass-stoppered bottles, and to determine after a lapse of time whether any alteration had occurred in their composition. Of the three samples taken, Sample A was kept for two years and examined at the end of each, and Samples B and C were only examined at the end of one year.They are interesting for several reasons. In the first place, they confirm Crispo’s work, as they show that there is a general tendency towards a slight increase in the Reichert-Meissl value. They also show that there is a similar tendency with the Polenske value to increase slightly. The differences observed, however, in the saponification and Valenta values are much more marked and definite.After one year’s keeping the saponification values increaaed respectively by 9.4, 7.6, and 8.4, and after two years the value for Sample The results are given in Table IV.250 TRIMEN: EGYPTIAN BUTTER AND SAMNA No. of Samples. A had increased by 16.3. The Valenta values show even greater differences-via., - 12O, - 12O, and - 11" C. respectively after one year, and after two years the Valenta, value of Sample A shows a decrease of 28' C.Reichert- Valen trt Zeiss (Acetic Acid : Saponification Meissl Polenske . Limits. at 40" C. gr. 1.060). Value. TABLE IV. Showing :he Efect of Ageing on the Composition of Samna. Sample. A.-Fresh ... ... A.-After one year A.-After two years B.-Fresh ... 0 . . B.-After one year C.-Fresh ...... C.-After one year Reichert-Meissl 31.6 32.7 33 6 32.3 33.4 31-9 31.6 Polenske. 1.9 2.5 3.1 2.0 2-3 1.6 1.6 Zeiss at 40" C. 42.0 42.0 42.0 42.1 42.5 42-2 43.0 Valenta (Acetic Acid sp. gr. 1.060). 58 45 30 59 47 62 51 Saponification Value. 230.9 240.3 247.2 230.5 238.1 227.7 236.1 ~- -_ Although few in number, these results certainly show that a considerable normal increase in the saponification value takes place during the ageingof samna, with a corresponding decrease in the Valenta value.I t is not proposed to discuss here the nature of the changes which take place during the ageing, but whatever they may be, they are most probably brought about by the constant high temperature ex- perienced in Egypt during the summer months. In any case, changes that would be only of academic interest in Europe and other temperate climates become of considerable practical importance in Eastern countries, where the analyst may frequently have to decide on genuineness of butter-fat some years after its manu- facture.Syrian Sumnu.-The mean results of a few samples of Syrian samna, believed to be genuine, are recorded in Table V., arranged according to their ascending Reichert-Meissl values : TABLE V.Mean Results from Syrian Samna, arranged according to the Reichert-Meissl Values. 24-25 25-26 27-2 29-30 30.6 31.2 26-27 28-29 4.3 4.8 5.8 5.3 6.7 6.7 7.9 6.4 42-6 42.5 41.9 43.0 42.2 42.4 41.5 42.3 1 65 60 - - 227.3 - - 232.8 - - - 235.5 I I I I ITRIMEN: EGYPTIAN BUTTER AND SAMNA 251 41.8 41.7 41.3 41.2 41.2 As remarked before, this kind of samna is generally believed to have its origin in the milk of sheep and goats, although some evidence has already been adduced to show that some samples at least consist entirely of sheep butter-fat.Attention has also been drawn to the extremely high normal Polenske values (4.3 to 6.4) and to the comparatively low Reicbert-Meissl values (24-6 to 31.2).I t will be noticed that the Polenske varies 2-1 for a corresponding variation in the Reichert-Meissl of 6.6. This variation is much larger than that found either in cow or buffalo butter-fat, and consequently the detection of small quantities of cocoanut oil in Syrian samna is a much more difficult matter than in their case. The difficulty is great when the samna is of known Syrian origin, but when its origin is unknown, the problem is at present practically without solution. Could one be certain that no goat butter-fat ever entered into the composition of Syrian samna, then the Hinks test could be employed with reasonable confidence ; but even that method is unsatisfactory as being only qualitative and difficult, if not impossible, of application in the hot weather. It was chiefly in the hope of finding a method that would solve this problem that the methods recounted above were tried. The solution, however, is still to be found, and it is hoped that additional work which is now being carried out will at least indicate the direction in which the solution lies. Indian Ghee.-The mean results of a few samples of Indian butter-fat arranged in t-t similar manner will be found in Table VI. 66.0 225.7 58.0 226.8 62.5 227.0 56.5 228.6 60.5 230.1 TABLE VI. Mean Results from Indian Ghee, arranged according to the Reichert-Meissl Values. I I I (Acetic Acid : Saponification Polenske. 1 Zeiss 1 Value. Reichert- Meissl Limita. at 400c* sp. gr. 1.060). No. of Semples . 2 28-29 1.7 3 29-30 1.6 2 30-31 1.6 2 31-32 1-9 1 32.8 1.6 The figures, which are very similar to those published by Revis and Bolton (ANALYST, 1910, 35, 343, and 1911, 36, 392), indicate that the samples examined consist essentially of buffalo butter-fat ; otherwise they do not call for further comment. In conclusion, the author wishes to lay stress on the fact that the paper is only of a preliminary character. He desires to express his thanks to Mr. A. Lucas for suggesting the investigation, and for his help and advice. THE SURVEY DEPARTMEKT LABORATORIES, CAIRO.

ISSN:0003-2654    

DOI:10.1039/AN9133800242

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

252 DROOP RICHMOND: NOTE ON SAMNA. THE COMPOSITION OF MILK. BY H. DROOP RICHMOND, F.T.C. (Rend at the Meeting, May 7 , 1913.) IN the present communication the results obtained during 1912 in the laboratory of the Aylesbury Dairy Company are summarised. The total number of analyses of milk was 39,179, and these are divided into several series. The largest and most important of these is that which refers to the milk as received from the farms, and the mean composition is given in Table I.The morning and eveuing milks are kept separate, and the numerical mean of the two milks is also given. This table, containing the results of 19,646 samples, is the most nearly representative of the true composition during 1912. It may be considered an objection that, as the morning and evening milkings differ somewhat in quantity, it is not strictly correct to take the numerical average, but I have found that the differences between the true average and that given are 80 small that it is only the third place of decimals per cent.that is affected. A confirmation of this is shown by the fact that another series of samples, representing the milk sent out from the chief and branch depots, differs only by very smallDROOP RICHMOND : COMPOSITION OF MILK 253 amounts in the monthly averages, and gives identically the same results for the yearly average ; and in this series, representing 11,677 samples, the differences, which may be due to varying proportions of the morning and evening meals, are completely eliminated.Three smaller series of samples, containing 1,480,2,341, and 1,903 samples respectively, also confirm the general average.TABLE I. Average Composition of Milk during 1912. Month. January ... February ... March ... April ... ... May ... .., June ... ... July ... ... August . . . September . , . October ... November ... December . Average ... Sp. Gr. 1 -0320 1 -0322 1 -0322 1 -0323 1 *0325 1 -0322 1-0317 1.0319 1 -0321 1 9323 1 -0321 1.0318 1.0321 Morning Milk.Evening Milk. Total Solids. % 12.37 12.30 12.19 12.20 12'10 12.15 12.20 12.22 12.45 12.67 12.93 12.70 12.37 Fat. - % 3-54 3'44 3.34 3.33 3-21 3-30 3.45 3'43 3.57 3.71 3.97 3-84 3-51 - ~ Solids not- Fat. % 8 *83 8-86 8-85 8.87 8 -89 8.85 8 -75 8-79 8-88 8-96 8-96 8.86 8.86 -~ Sp. Gr. 1.0318 1.0318 1 -031 7 1.0320 1 -0320 1.0318 1.0312 1.0315 1.0318 1,0321 1-0319 1.0316 1.0318 Total Solids.% 12.59 12.55 12-45 12-48 12-52 12-60 12-55 12-75 12.88 13-03 13.18 12.86 12.70 Fat. % 3.76 3.71 3-62 3'62 3.66 3.76 3-84 3.95 3.99 4-06 4-23 4-02 3-85 Solids. not- Fat. % 8-83 8.84 8.83 8-86 8-86 8.84 8.71 8.80 8.89 8.97 8.95 8-84 8 *85 Sp. Gr. 1.0319 1,0320 1 -0320 1.0321 1 -0322 1 -0320 1-0315 1-0317 1 -0320 1 *0322 1 -0320 1.0317 1.0319 Mean. Total Solids - % 12.48 12'42 12.32 12-34 12-31 12.37 12.37 12-48 12.66 12.85 13-05 12-78 12.54 Fat.- % 3.65 3 -57 3-48 3 '48 3.43 3.53 3.64 3-69 3.78 3.89 4-10 3.93 - 3 -68 Solids- not- Fat. % 8-83 8.85 8-84 8-86 8.88 8-84 8.73 8.79 8.88 8-96 8-95 8-85 8-86 The average percentage of fat (3.68) is lower than that found last year, and, indeed, is the lowest noted since 1900.This was due to the percentage of fat being lower than usual during the first five months of the year. As a rule it has been noticed that May and June are the two months in which the lowest percentages of fat are found, but during 1912 the months of March, April, and May were thcse in which the milk was poorest, and there was a distinct improvement during June.I attribute the low percentage of fat during the early months of the year to the effect of the prolonged drought of 1911 on the quantity and quality of the feeding- stuffs. I t is probable that the rise in the amount of fat noticed in June is to be attributed to the unusually cold summer. As is usual, the highest percentage of fat is noticed in November, but the figure found (4.10) is higher than has been found for many years. The average difference between the morning and evening milks (0.34 per cent.) is a normal one.During 1912 there was the usual fall in solids-not-fat during July and August, but very few cases of figures falling below 8.5 per cent. were found, and in Septem- ber the solids-not-fat were normal. There was no repetition of the experience of 1911, when low solids-not-fat were frequently found from July to October.In continuation of the tables given in former years, I give the percentages of the morning inilks falling below 3.0 per cent. of fat. As a rule it is only from April254 DROOP RICHMOND : COMPOSITION OF MILK to June that any appreciable number of samples containing less than 3-0 per cent.of fat is noticed ; but in 1912 there were samples low in fat during the months of January fo August inclusive, and consequently the table is much extended. In the evening milk and in the morning milk during the other months of the year samples yielding below 3.0 per cent. of fat were very few in number. TABLE 11. I _- Month. 1 2.9 t o 3.0. January ... ... ... February ... ... ...March ... ... ... April ... ... ... May ... *.. ... June ... ... ... July ... ... August ... ... ... ... 0.4 1.2 1.8 1-1 4.6 3.9 0.7 0.8 2.8 to 2.9. 0.4 0.5 0.6 0.9 4.2 0.9 0.2 0.2 2 7 to 2.8. 0.2 0.5 0.5 0.6 1.6 0.7 - - Below 2.7. 0.5 0.3 0.3 0.3 0.4 0.5 0.2 - During 1912 a systematic determination of the aldehyde figure, from which the percentage of proteins was calculated, and of the ash, was made, and the results of these, as well as the percentages of milk-sugar (by difference) appear below.TABLE 111. Month. January _.. ... ... February ... ... ... March ... ... ... April ... ... ... May ... ... ... June ... ... ... July ... ... ... August ... ... ... September.. . ... ... October . , . ... ... November . . . ... ... December . . . ... ... Solids-hot-Fat.8.83 8.85 8-84 8-06 8.88 8.84 8-73 8-79 8*88 8-96 8.95 8-85 Milk Sugar. 4.58 4.67 4.70 4.69 4-61 4.63 4.57 4-56 4.52 4.60 4-64 4.59 Proteins. 3-50 3.42 3.40 3.43 3.54 3-48 3.42 3.48 3.60 3.60 3.55 3.50 Ash. 0.75 0.76 0.74 0-74 0-74 0.73 0.74 0 -75 0.76 0.76 0.76 0.76 I t is seen that, as was noticed in 1911, the percentage of proteins tends to follow that of the solids-not-fat, though the variations during 1912 are very much less.I wish to acknowledge the great assistance given me by Mr. H. C. Huish, to whom the bulk of the above figures are due. DISCUSSION. Mr. J. GOLDING said that he had been asked whether any information was available as to what was the breed of the cows producing these samples. He thoughtDBOOP RICHMOND : COMPOSITION OF MILK 255 that the value of the tables would be greatly increased if such information could be given, and also if, in the event of there being any herds which included, for example, Jersey cows, the figures apart from these could be given.l&. C. REVIS said that he was afraid it would be rather difficult to obtain exact information as to the breed of the cows. His own average figure for fat in 1912 was 3.65, Mr. Richmond’s average being 3.68; and hio experience was the same as that of Mr. Richmond as to the months in which the fat was lowest and highest, his highest figure, as in Mr.Richmond’s case, being obtained in November, for which month his average was 4.05. Such an agreement was, indeed, only to be expected when the bulks of milk dealt with were so large; but it was only fair to point out that such averages ought not to be used in judging the milk from individual farms, where, quite apart from any question of adulteration, the composition of the milk might vary considerably, according to the breed of the cows.His averages were, as he had explained last year, obtained in a somewhat different way from Mr. Richmond’s, being based on analyses of samples representing large bulks of mixed milk, instead of on samples representing each consignment individually. I&. W. PARTRIDGE asked whether the samples examined by Mr. Richmond and those examined by Mr. Revis came from the same counties. Mr. REVIS said that the two sets of samples came largely from the same areas, which, however, were in both cases very large, extending over the whole of the Southern Counties and as far north as Oxfordshire-perhaps even to Cheshire. The figures, however, would not apply to the Northern or to the Eastern Counties.

ISSN:0003-2654    

DOI:10.1039/AN913380252b

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

DBOOP RICHMOND : COMPOSITION OF MILK 255 THE PROPORTIONATE DETERMINATION OF COCOANUT AND PALM- KERNEL OIL IN MIXTURES. BY H. R. BURNETT AND CECIL REVIS. (Read at the Meeting, May 7, 1913.) AT the present moment there is no doubt that cocoanut and palm-kernel oils form the main constituents of many margarine mixtures, and for this reason some means of arriving at the relative proportions of the two fats is very necessary to the technical chemist.There is no great difficulty so long as a (‘ straight” mixture of the two is being dealt with, but as in practically all cases some other fat, usually a liquid oil, is present with them, the problem is a little more complex. There is probably no exact chemical means by which a solution may be arrived at, on account of the great similarity between these two fats; and, indeed, in all cases the result must be really arrived at by mathematical juggling with the ascertained figures and the average values which the oils give.Since these values vary between somewhat wide limits, it follows that the results are approximate, whilst, if any error be made in the determinations, the final calculation may be very far from the truth.In a256 BURNETT AND REVIS : PROPORTIONATE DETERMINATION OF paper published last year (Bolton, Richmond, and Revis, ANALYST, 1912, 37, 183), it was shown that either fat alone could be estimated with a high degree of accuracy by means of the Reichert-Meissl-Polenske method, if carried out in a standard way, the Polenske figure being the real determining factor ; but the similarity in character of the curves for cocoanut and palm-kernel oils preclude any exact knowledge of the quantities present when both are exhibited together, unless some exact idea of the quantity of the other constituents present be known.If this can be estimated, the Polenske figure determines the mixture by simple interpolation ; but as usually some liquid oil, such as cottonseed, etc., has been used, for which no exact method of estimation is available, the Polenske figure does not give the necessary information.In these cases it is only possible to fall back on the saponification value, and to arrive at the result by interpolation, assuming a certain value (which will usually be correct) for the third constituent of the mixture, when the Polenske figure becomes available as confirmatory evidence.As this process depends on the assump- tion that the saponification value taken for the third constituent is correct, we have devised the following simple addition to the Reichert-Meissl-Polenske process, not as an exact method of analysis, but as providing a useful check on the mathe- matically deduced result.We believe, however, that the figures obtained by this method are within the limits of error of such results. In an ordinary “straight ” mixture of cocoanut and palm-kernel oils, the Polenske figure will determine the proportion with at least as great exactitude as any other method. For instance, if Polenske values are plotted as abscissae, with percentage composition as ordinates, then a straight line joining the points which represent 100 per cent.cocoanut oil and 100 per cent. palm-kernel oil respectively will include the Polenske values for all mixtures of these two. The following process is for mixtures containing other constituents : The ordinary Reichert-Meissl-Polenske determination is made in the standard apparatus and by the standard method. The Polenske figure is obtained using FG baryta.The insoluble barium salts are then filteredoff on a hardened filter-paper under pressure, and the salts washed three times with 3 C.C. of 93 per cent. alcohol (by vol.), the funnel being kept covered during filtration and washing. The paper, after all possible alcohol has been sucked out, is dropped into a wide-mouthed CO, flask, ten times the Polenske value in c.c.’s of 93 per cent.alcohol* (by vol.) added, and the flask boiled under a reflux condenser till the barium salts are in solution. About 5 C.C. of the hot solution are then poured rapidly into a strong test- tube (6 inches x & inch), which is at once closed with a stopper carrying a small bulb thermometer and an aluminium wire stirrer. The liquid is rapidly stirred, holding the tube in a good light, and the turbidity-point noticed.The liquid is then warmed till again clear, and the turbidity-point again noted. This second temperature is taken as the turbidity temperature. If desired, the tube can be fixed in a wider tube, so as to obtain slower cooling. * The alcohol used in these experiments had a sp. gr. of 0.8235 at ’55”’ 155°C’ Alcohol of the right strength may be obtained by placing 7 C.C. of water in a 100 C.C.flask and making up to the iiiark at 18’5” C. with Kahlbaum’s absolute alcohol.COCOANUT AND PALM-KERNEL OIL IN MIXTURES 257 Working in this manner, cocoanut oil gives a turbidity temperature of 52.5' C. and palm-kernel oil 68.5" C., and mixtures of these fats give temperatures between these limits proportionate to the percentage composition, The turbidity-point is very sharp, is independent of the outside temperature, and the barium salts, on which the test depends, are quite insoluble in the cold alcohol used for the Polenske deter- mination.The turbidity-points are also quite independent of the amounts of the two fats present in the original sample, but determine their relative percentages, and SO supply the necessary information.The strength of alcohol (93 per cent. by volume) must be strictly adhered to if the values here given are to be employed. I t is the most satisfactory concentration. Other oils and fats (such as are likely to be present) do not interfere. In certain cases small quantities of insoluble volatile acids distil in the Polenske method, which give barium salts insoluble in boiling 93 per cent.alcohol. The turbid liquid is therefore poured into a long test-tube, corked, and kept upright in a water-bath at 70° to 71" C. until the solid matter has settled. The clear supernatant liquid is then poured off into the turbidity-tube and the temperature of turbidity determined. This process does not affect the results. This permanent turbidity, due to barium salts of acids other than those derived from cocoanut and palm-kernel oils, must be distinguished carefully from that due to palm-kernel " stearine." The barium salts of the insoluble volatile acids of this " stearine " do not dissolve in ten times the Polenske value in cubic centimetres of 93 per cent.alcohol, but the liquid becomes more turbid immediateZy the flask is removed from the water-bath.B s long as mixtures of cocoanut oil and palm-kernel oil are being dealt with, the above method gives good confirmatory evidence of the relative percentages. We have found that cocoanut oils of different Polenske values give practically identical turbidity temperatures. Samples of palm-kernel oil are not easy to obtain, and we cannot say from actual experience whether they will always give the same turbidity temperature as we have found, but probably the divergence will be small.It is to be noted that the filtration of the barium salts and their subsequent solution must be carried out within, a few hours of the Polenske titration, as otherwise the salts become partially insoluble and the results are inaccurate. Mixtures of the '' oleines " and of the " stearines " are sometimes employed in place of the whole oils, but probably rarely.I n these cases, though the turbidity temperatures do not give such accurate information, on account of the very variable composition of these products, at the same time they give most useful information as to their presence.From samples which we have obtained, we have found that palm-kernel Palm-kernel " stearine " gives barium salts which only partially go into solution in boiling 93 per cent. alcohol, and it is necessary to make dilutions to obtain the true turbidity tem- perature. As dilutions with one-third and one-half the volume respectively, gave temperatures of 63.5" and 54.5" C., it follows that the true temperature for this sample of stearine " (saponification value 247.6) was 72.5" C.Samples of cocoanut '' oleine " and '( stearine" could not be obtained, but from mixtures with palm-kernel ( ( oleine " and L c stearine '' it was calculated that, in In such cases a clear solution cannot be obtained. . oleine " (sap. value 244.8) has a turbidity temperature of 59.5" C.258 COCOANUT AND PALM-KERNEL OIL IN MIXTURES the samples examined, cocoanut ‘‘ oleine ” gives a figure of 53” C.and the ‘‘ stearine ” 63’ C. These figures, taken in conjunction with the Reichert-Meissl and Polenske figures, and the iodine value of the sample, will make it quite clear whether the whole oils or their products are present. The following table gives the results obtained : Fat.Cocoanut oil ... ... ... ... Palm-kernel oil . . . ... ... ... Palm-kernel ‘ 6 oleine ” . . . ... ... Palm-kernel ‘( stearine ” ... ... Cocoanut ‘ I oleine ” ... ... ... Cocoanut I ‘ stearine ” . . . ... ... Cocoanut ‘( oleine,” 80 per cent. I Palm-kernel “ oleine,” 20 per cent. j “. Palm-kernel “ stearine,” 40 per cent. Cocoanut ‘( stearine,” 60 per cent.Reich ert- Meissl. 7.5 5.2 7.2 - - - 8.24 4.43 16.5 9-6 12.1 8.2 - - 17-05 9.93 Turbidity Temperature. * c. 52.5 68.5 59.5 72.5 53.0 (calc.) 63.0 ,, 54.5 67.0 It is interesting to note that cohune oil, which is identical analytically with In conclusion, we desire to thank Mr. E. R. Bolton for a valuable suggestion cocoanut oil, gives exactly the same turbidity figures. during the course of the work.DISCUSSION. Mr. E. R. BOLTON said that the authors had kindly given him opportunities of trying this method while their work was in progress, and he had found the two oils to be distinguishable by the difference indicated in the paper. According to his experience, however, it was necessary to be very careful that the alcohol was of the proper strength, and he thought it desirable to obtain the alcohol in large batches, and to make a blank determination for each. The case appeared to be like that of the Valenta test or the Maumen6 test, in which the results would be affected by differences in the strength of the reagents, too small to be within the limibs of analysis. The PRESIDENT suggested that it might be better to indicate the strength of the alcohol by stating its specific gravity, since confusion might possibly arise between percentage by weight and percentage by volume. Mr. REVIS, in reply, said that the strength of the alcohol need not necessarily be 93 per cent., but it should not be greater than that, or the barium salts which palm-kernel oil yielded would not go into solution. If weaker alcohol were used, the results would still be comparable, but the widest range of temperatures was obtained when the strength was 93 per cent. by volume.

ISSN:0003-2654    

DOI:10.1039/AN9133800255

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

EXAMINATION OF THE OILS FROM MANIHOT CEARA, ETC. 259 EXAMINATION OF THE OILS FROM MANIHOT CEARA AND FUNTUMIA ELASTICA AND A COMPARISON OF THEIR PROPERTIES WITH THOSE OF LINSEED AND HEVEA OILS. BY S. RIDEAL, D.Sc. (LoND.), F.I.C., AND L. H. D. ACLAND, B.A. (OXON.). (Read a t the Meeting, May 7, 1913.) Oil from Ceara Rubber Seeds.-The seeds were about the size of a small hazel nut, had a very hard shell, and were covered with a reddish powder, which, judging from the high ash and silica content of the cake, was probably clay.The seeds are difficult to work with, not only on account of the hardness of the shells, but also because the kernels stick to the shells, making the seeds diEcult to decorticate. A fairly good separation may be made by pounding the seeds in an iron mortar or passing them through a mincing- machine, and then separating them with a piece of coarse gauze through which the kernels pass.This method gave 45.3 per cent. kernel and 54.7 per cent. of shell. The separation is not complete, and the kernel contains a certain amount of sharp pieces of shell, rendering the oil-cake unfit for food unless very finely ground. The oil is light yellow-particularly light when extracted with petroleum ether, or pressed-with a faint green fluorescence.It has a very agreeable nutty odour and taste. The content of oil was, in the kerneI, 35 per cent.; in the whole seed, 15-75 per cent. The only specific test for the unrefined oil noted was the fact that when it is heated with about 1 to 2 per cent. of manganese resinate, it turns a bright pink colour.I t resembles linseed oil in several respects. It throws down similar mucil- aginous matter on standing ; has a high iodine value, dries fairly quickly, giving a tough and distinctly elastic skin which is very white in colour; in fact, it is an excellent drying oil. When boiled, it behaves like a sample of freshly extracted linseed oil in that it becomes very viscous.It does not get very dark under these circumstances. Treated with sulphur chloride, it gave at once a transparent gelatinous mass. On warming this, it darkens and gets rather harder, forming a fairly elastic mass. Five grms. of the oil gave 6.16 grms. of vulcanised oil. Oil from the Seeds of Funtumia E1astica.-The seeds are small and pointed at the ends, and are roughly crescent-shaped.The husks are very thin and soft, so that decortication is unnecessary. The weight of 100 seeds is 4.8 grms. The oil content is 31 to 33 per cent. on the whole seed. The oil is very dark in colour. It was a t least dark yellow, however it was extracted; and when extracted with trichlorethylene, it was dark brown. It has a slight green fluorescence.The oil, extracted with trichlorethylene, is very difficult to decolourise. Sulphurous acid, nitric acid, and hydrogen peroxide do not affect its colour. A few drops of sulphuric acid (1 : 3) did not decolourise it appreciably. Chlorine had a slight decolourising efiect. The best method is to heat the oil to about 100" to 120" C. for several hours with a mixture of fullers' earth (dried by exposure to the air, as 3 per cent.of water must be present in it for it to The weight of 100 seeds was 53-2 grms.260 RIDEAL AND ACLAND: EXAMINATION OF THE OILS be effectual), mixed with recently ignited animal charcoal. The oil has a strong, bitter odour and taste, and also froths rather on boiling.” The oil deposits a little mucilaginous matter on standing; but later, after about ten days, it deposits a quantity of stearine in cold weather.I t has the smell of the original oil, and gives a m.-p. of 4 7 O C., melting to a clear brown liquid. Its saponification value is 1844. The oil is a drying oil, but not a very good one, as it gives rather a tacky ” film. When heated with 0.02 per cent. of manganese resinate, it became slightly gelatinised, but not sufficiently so to prevent it from flowing.When boiled it became rather dark in colour, and formed a very viscous liquid resembling a, sample of linseed oil (freshly extracted from linseed meal) similarly treated. Heated with sulphur chloride it gave at once a brown gelatinous mass. On warming this to drive off the carbon bisulphide, it darkened and formed a very elastic but sticky substance.Five grms. of the oil yielded 6.6 grms. of substitute. The oil is very suitable for making rubber substitute. I t s bitter taste and rather high dextrorotation- + 0.7, whereas the other oils of the linseed group are all laevorotatory-suggested the presence of an alkaloid. Dragendorff’s method was used in testing for the alkaloid. The extract was 0.11” per cent.of the oil taken, and was extremely bitter. This gave a yellow precipitate with potassium iodide, and with picric acid a yellow precipitate insoluble in ammonia. The alkaloid was volatile, as the water in the distillate when the trichlor- ethylene was being blown out of the marc with steam was distinctly bitter. This alkaloid is evidently very unstable as, on extracting the washed seeds with slightly acidulated water, a, very bitter solution is obtained. This solution, however, lost its bit’terness on warming with lead acetate and basic lead acetate (this mixture does not precipitate any known alkaloid). This might have precipitated a glucoside, but the solution contained no reducing sugar, and glucosides are insoluble in oils ; there- fore the bitterness of the oil cannot be due to a glucoside.The seeds contain a cyanogenetic glucoside in rather a large quantity, as on adding a little very dilute acid (not more than 1 per cent.), not only was a precipitate produced in a drop of silver nitrate, but also the smell of prussic acid was very noticeable. Nethods of Analysis.-The iodine values were all done by Wij’s method, allowing the flasks to stand for not less than forty-five minutes, and leaving as much iodine over as was absorbed.In the Hehner value the acidification was performed with hydrochloric instead of sulphuric acid, as potassium chloride is more soluble than the sulphate. The relative viscosity was determined at 15” C. by running the oils out of a burette, and the times for the oils compared with the time taken by water to run from the same burette.I n determining the other values standard methods as given by Bolton and Revis in their book ‘‘ Fatty Foods,” or in the case of methods not given by them those described by Lewkowitsch, were used. The figures in the comparative table below were taken from the same sources. The method of S. Fachini and W. Dorta (Chern.Rev. Fett.-Ind., 1912, 19, 77) for the separation of the solid and liquid acids by the solubility of the potassium salts of The stearine is brownish-white, and rather soft and sticky.FROM MANIHOT CEARA AND FUNTUMIA ELASTICA 261 the liquid acids in 90 per cent. acetone was not successful. Five grms. of each of the oils were treated as in the lead-salts-ether separation, and the potash soaps extracted at the ordinary temperature with three portions of 50 C.C.each of 90 per cent. acetone. The acetone was boiled off and the soaps taken up with water, the acids liberated, and then treated with ether until the aqueous solution was clear. This method gave- Acetone Method. Lead-salts-Ether Method. Per Cent. Per Cent. Funtumia (liquid) . . . ... 28.4 79.8 Ceara (liquid) ...... 39.1 88.9 I n both the lead-salts-ether and potash-salts-acetone methods for separating the saturated and the unsaturated acids, the liquid--i.e., unsaturated, acids were freed from ether by distillation in an atmosphere of carbon dioxide under reduced pressure. In this way the acids are obtained in a very pure state, being neither injured by heat nor discoloured by oxidation.The process is very simple and very rapid provided that a sufficiently fine capillary tube is used to conduct the carbon dioxide into the distilling flask. In finding the Reichert-Meissl value, it was noticed that Funtumia oil gave no solid volatile but insoluble acids, whereas Ceara, oil gives a large quantity of them. Both of these oils are very like Hevea oil in all their properties.Not only are all three good drying oils, but also all their values are close together, as will be seen from the comparative table given below. Ceara oil differs from Hevea oil in that the oil-cake contains no cyanogenetic glucosides, whereas Funtumia oil does contain one. Both differ from Hevea oil in that they do not contain an appreciable amount of lipolytic enzymes.Physical and chemical constants of the oils compared with linseed and Hevea oils : Weight of 100 seeds ... Oil content of kernel ... Oil content of total seed ... Sp. gr. ... ... Refractive index (15; C.) ... Relative viscosity . . . ... Iodine value ... ... Saponification value . . . Acid value . . . ... ... Hehner value ... ... Reichert-Meissl . . . ... Iodine value thereof .. . Liquid fatty acids ... ... Ceara . 53-2 grms. 35.0% 15.75% 0.9238 1.475 13.0 135.0-137*0 189.1 0.625 95.1 0-44 88.9% 162.5 Funtumia. 4.8 grrns. 31*0-33*0% 0.9320 1.4788 - 14.3 138.0 185.0 2.65 94.0 0.66 79.8% 175.5 Fat -Free Residue. Per Cent. Nitrogen ... ... 2-19 Proteins .,. ... 14.23 Ash ... ... ... 17-10 Sand ... ... ... 0.72 Potash ... ... 2.65 Phosphoric acid ...1.72 Hevea. 360 grffis. 45.48% 22.25% 0.9258 - - 138-7 191.2 5.26 2.7 86.0% - - Per Cent. 4-34 27-08 5.04 0.38 1.44 2.25 Linseed. - - 36.40% 0.931-0.938 1.4835 18.0 19 2.0-200 SO below 5.0 94081-95.5 0.00 92.5% 190.0-195.0 -262 RIDEAL AND ACLAND: EXAMINATION OF THE OILS Other figures are given for the value of these oils. The figures for Ceara oil given by Fendler and Kuhn (Ber. deut.Phurm. Ges., 1906, 15, 426), quoted in the BuEZetzn, of the Imperial Institute, 1906, p. 364, are very close to those given by US. The slight differences between these figures and ours would be accounted for by a difference in the source of the seeds; the seeds examined by us came from British East Africa. A. Hebert (Caoutch. et Guttapercha, 1912, 9, 6232-6234) quotes the following remarkable figures for funtumia oil : Oil content ...... ... ... Sp, gr. ... ... ... ... ... Acid value ... ... ... ... Saponification value . . . ... ... Hehner value . . . ... ... ... Reichert-Meissl value . . . ... ... Iodine value . . . ... ... ... Liquid acids ... ... ... ... 20.0% 0.929 1.8 167.6 94.0 12.1 139.0 60% None of the glycerides occurring in the natural oils, especially the drying oils, have a saponification value below 170.An oil with a Reichert value of 12.1 would naturally contain a large proportion of the glycerides of the lower fatty acids, and in this case one would expect a very high saponification value-e.g., cocoanut oil with 8 Reichert of 7.5 has a saponification value of 257. Such a large difference, a difference in kind and not only in degree, could not be accounted for by any variation in the environment in which the trees were grown.Even if his oil were not from Funtumia elastics, it is extremely difficult to see how his values for the Reichert and saponification values could be reconciled. We desire to thank Mr. E. R. Bolton for his kindness in extracting samples of the seeds for us.DISCUSSIOK. Mr. R. G. PELLY said that some years ago he had occasion to examine a sample of Ceara rubber seed, with results similar to those obtained by the authors, and had come to the conclusion that it would not be practicable to use it commercially. The oil only amounted to 15 per cent. of the whole seed, and as the shells were very hard, and could not be removed without breaking up the kernels, the only practicable plan seemed to be to crush the whole seed and extract the oil by a, solvent. The residual meal from the whole seeds, which was poor in proteins, could hardly be used as a cattle-food owing to the large amount of woody matter from the shells, would not fetch more than 64 per ton as a fertiliser, and, the yield of oil being so small, it seemed hardly likely that it would pay to export the seeds, although there was a fairly large supply of them.On the other hand, the use of extraction apparatus in the tropics, worked by native labour, did not seem to afford much promise of success. Mr. L. GOWINU SCOPES said he had found it necessary first to dry the seed or other material in order to extract the oil completely with trichlorethylene,.although some workers in Germany appeared to have obtained satisfactory results without previous drying. It might be mentioned that practically all the trichlorethylene nowFROM MANIHOT CEARA AND FUNTUMIA ELASTICA 263 available contained a small proportion of hexachlorethane, which it was almost impossible to remove, and which always remained behind with the fat.A method, however, was at present under investigation, based on the formation of a compound between the trichlorethylene and mercuric cyanide, by which it was hoped to prepare the solvent free from hexachlorethane. Mr. CHASTON CHAPMAN remarked that if, in the case of extracted drying oils, the oil were not heated to a sufficient extent to remove the solvent completely, such traces as remained might affect the specific gravity and refractive index of the oil ; whilst, on the other hand, a degree of heating sufficient for the removal of the solvent might affect the properties of the oil in other directions.He should be interested to hear whether these two oils resembled linseed oil in yielding brominated glycerides insoluble in ether. Mr. HEHNER said that in the case of cocoanut oil extracted with carbon bisulphide he had found it impossible to remove the solvent completely by steaming.The PRESIDENT said that at one time he used often to find, in examining rape oil extracted by petroleum ether, that a sufficient quantity of the solvent remained behind to lower the flashing-point of the oil very considerably, such oil often having a flashing-point not greatly above 200” F., whereas the normal flashing-point of rape oil was above 400” F.; and on being heated in the water-oven the oil would lose fr6m 0.3 to 0.5 per cent.in weight. When this was pointed out to the manufacturers, they took steps to effect the more complete removal of the solvent, with the result that the flashing-point of these extracted oils was now generally normal.With regard to the “ viscosity ” determinations, he would like to suggest that the results should be called what he presumed they were-viz., “efflux times,” the kind of viscosimeter used and the temperature being stated. The comparative figures which the authors had given for linseed oil struck him as being rather peculiar. The range of iodine value, for instance, was given as 192 to 200; but in the case of raw linseed oil it was rarely so high as 200, while lower figures than 192 were frequent.The saponification value mentioned-namely, 195-also struck him as rather high for linseed oil. For the extraction of oils in the laboratory from paints, etc., he always preferred to use ether, because this was so readily expelled, whereas petroleum spirit, unless freshly redistilled, was liable to contain substances of a less volatile character.In extracting the oil from paint, if the last traces of ether were driven off in an atmosphere of carbon dioxide, the oil was obtained in such a, condition that its specific gravity, iodine value, and yield of bromides, were approxi- mately normal if genuine linseed oil had been used.Dr. RIDEAL, referring to the question of commercial possibilities, said that both these plants were grown to a very large extent in German East Africa, and German West Africa. The funtumia seeds which they had examined came from the Cameroons, and the ceara from British East Africa; but ceara rubber was being produced on a large number of estates in German East Africa, and the quantity of seed produced on both the East and the West Coast would be very considerable in the near future.In the Malay Peninsula and Ceylon hevea seed had been largely used for fresh planting during the last few years, so that hevea oil had not yet become a commercial article; but of course there was a limit to the amount of264 ABSTRACTS OF CHEMICAL PAPERS planting that could take place.It was not likely that much more funtumia would be planted in West Africa, as hevea was beginning to be introduced there. Ceara grew very quickly, and yielded a Iarge quantity of seed. The seed was certainly very hard, and contained only about 15 per cent. of oil; but he thought it would be possible in East Africa, in both the English and the German colonies, to extract the oil locally and use the residue as manure.Mr. ACLAND said that they had made a few experiments with driers, the results of which were not yet published, using lead, manganese, cobalt, uranium, and vanadium linolates and resinates, and also cobalt borate, acetate, and benzoate, these last three being insoluble. They had found the drying properties of the ceara oil to be fairly good, not far removed from those of linseed oil. Funfumia, oil was not quite so good as regards drying. Soya oil came next, and finally hevea oil. They had used petroleum ether in preference to ordinary ether because the latter seemed to dissolve a good deal of colouring matter, which was not the case with petroleum ether, and provided that tha petroleum ether was of suitable quality, boiling at about 40" C., they had found that it was an excellent solvent for the purpose. The heating of drying oils did not do much harm provided there was no air present. It was quite simple to drive off most of the petroleum ether in a current of carbon dioxide, and the iodine value did not appear to be altered if the filtering was done through a folded paper in the oven. It was only in the cme of a thin film of oil that oxygen would be absorbed quickly enough to have an appreciable effect. He did not say that filtering in the drying oven would have absolutely no effect, but he thought that any such effect would be within the ordinary errors of analysis. The oils yielded no appreciable quantity of hexabromide.

ISSN:0003-2654    

DOI:10.1039/AN9133800259

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

264 ABSTRACTS OF CHEMICAL PAPERS ABSTRACTS OF PAPERS PUBLISHED IN OTHER JOURNALS. FOOD AND DRUGS ANALYSIS. Adulteration of Barley Meal. Sendhoff and Weinstein. (Chem. Zeit., 1913, 37, 485-487.)-Among the adulterants used in German barley meal are milling offals, consisting of all the foreign seeds and refuse sorted out at the mills ; barley bran, consisting of the hulls and points of the corns separated in the preparation of pearl barley and similar products ; ‘‘ English dust,” consisting mainly of chalk, together with refuse from all kinds of cereals, foreign seeds, coffee berries, etc.; (‘ F meal,” consisting of about 70 per cent. of calcium sulphate and 30 per cent. of sawdust ; rice bran ; ground maize-ear refuse ; ivory-nut meal ; tapioca meal. Preliminary tests may be made by sedimentation in chloroform, in which mineral impurities fall to the bottom.Chemical tests incrude determinations of total ash, mineral matter insoluble in hot 10 per cent. hydrochloric acid, and crude fibre by Konig’s glycerol-sulphuric acid method. For the microscopical examination, about 2 grms. of the meal areFOOD AND DRUGS ANALYSIS 265 treated in a basin with 2 to 3 C.C.of fuming nitric acid and 100 C.C. of water, and boiled for about ten minutes. After settling, the liquid is poured off and the insoluble matter collected on a filter, When this residue is treated with chloral hydrate solution (1 : l), clear microscopic preparations of husks and foreign particles are easily obtained. Normal barley meal shows the following composition : Mineral matters, 2.5 to 3.5 per cent.; matter insoluble in hydrochloric acid, 0-5 to 1-0 per cent. (sand, etc.) ; crude fibre, 4.0'to 5.0 per cent. ; sulphuric acid, traces ; calcium oxide, 0.5 per cent. Similar average analyses are cited for the principal adulterants enumer- ated above, and from these, taking crude fibre as a basis, the approximate proportions of any microscopically ascertained adulterant may be calculated.J. F. B. Reactions of Digitalis Glucosides. Digitonin. C. Reichard. (Pharm. Zentralbl., 1913, 54, 217-221 ; through Chem. Zentralbl., 1913, I., 1368-1369.)--Digi- tonin, a glucoside found in digitalis leaves, has no physiological action on the heart, and has not been used in medicine; it combines with cholesterol to form a choles- teride, the reaction being quantitative.I t is insoluble in water, ether, and chloro- form, slightly soluble in alcohol, and dissolves in concentrated sulphuric acid to form a colourless solution which turns brown when heated, or after long exposure to the air. With cold 25 per cent, nitric acid digitonin yields a colourless solution which becomes yellow when heated: crystals form when this solution is allowed to evaporate spontaneously.A dirty, greenish-black coloration is obtained when digitonin is heated with potassium dichromate and sulphuric acid ; with ammonium molybdate and sulphuric acid a light blue coloration is produced, the colour changing after a time to blue-black. Digitonin yields a green coloration, changing to violet when heated with vanadic acid, a violet-black coloration with iodic acid, and a black- green colour with tungstic acid, A solution of digitonin in acetic acid decolourises bromine water, and reduces silver and mercury salts.A characteristic reaction of the glucoside is as follows: A drop of concentrated cobalt nitrate solution is evaporated on a microscope slide until it turns deep blue in colour, when a small quantity of solid digitonin and a drop of glacial acetic acid are added; after the lapse or from twenty-four to thirty-six hours, the mixture is converted into a mass of red- coloured hexagonal crystals.w. P. s. Composition of Honey. J. Fie he and P. Stegmuller. (Arbeit. Kaised. Gesundheitsamte, 1913, 44, 78-149.)--Bnalyses of 112 samples of honey are recorded, the samples representing the honey imported into Germany from Austria (nine samples), Hungary (four), Russia and Poland (twenty-seven), Italy (ten), Spain (seventeen), Portugal (two), North America (six), Mexico (three), South America (twenty-three), Cuba (four), Jamaica (three), and Australia (three).The quantity of water in the samples varied from 14.94 to 24-28 per cent., the average being 18-30 per cent.; in only three cases did the quantity exceed 28 per cent.The invert sugar ranged from 61.96 to 78-84 per cent., the average amount being 73.48 per cent. ; 108 of the samples were laevorotatory. Four of the samples contained upwards of 10 per cent. of sucrose, the maximum quantity found being 15.40 per cent. ; the least quantity found was 0.12, and the average 2-42 per cent.The acidity varied from 0.027 to 0,204 per cent.,266 ABSTRACTS OF CHEMICAL PAPERS expressed as formic acid ; the ash from 0.027 to 0.673, with an average of 0.150 per cent., and the phosphoric acid (as PO,) from 0.0075 to 0.0932, the average being 0.0198 per cent. Only one sample yielded a positive reaction with Fiehe’s test for invert sugar, whilst eighteen samplq yielded a positive reaction with Ley’s test ; the presence of starch syrup was not detected in any o€ the samples examined, and of ninety-five samples tested for invertase eighty-seven yielded a positive reaction.Lund’s tannin precipitation method was found to be of little use in ascertaining the purity of the samples. (See also ANALYST, 1912, 38, 449). w. P. s.Estimation of Proteins in Milk by Aldehyde Titration. W. C. de Graaff and A. Sehaap. (Ann. Falsif., 1913, 6, 149-157.)-The aldehyde method described by Steinegger (ANALYST, 1906, 31, 45) for the estimation of proteins in milk was found by the authors to be useful and trustworthy. sodium hydroxide solution for the titrations, having found that under the conditions of the method the dilution is not such as to augment the acidity of the formaldehyde (cf., Richmond and Miller, ANALYST, 1906, 31, 224; 1911, 36, 9).They deduce the factor 0.0777 for converting the aldehyde figure into percentage of nitrogen, and 0,495 for converting it into percentage of protein. These factors agree almost exactly with those given by Steinegger, but differ slightly from the factors given by Richmond (loc. cit.).In the case of human milk, the nitrogen factor was found to be 0,0693, and the protein factor 0.443. Butter-milk (six samples) yields aldehyde figures varying from 10.41 to 11-29 ; compared with the nitrogen found by Kjeldahl’s process, these figures corresponded with a nitrogen factor of 0.0498 and a protein factor of 0.318. Results of further experiments showed that it is not the formation of lactic acid which causes sour milk to have a high aldehyde figure, but that the increase is They employed due to the action of micro-organisms on the proteins present.w. P. s. Estimation of the Bromine Absorption of Wines. T. von Fellenberg. (Mitt. Lebensmitteluntersz~ch. Hyg., 1912, 3, 97-99 ; 1913, 4, 14-41 ; through Chem Zentralbl., 1912, II., 554 ; 1913, I., 1367-1368.)-Details of a process are given for the estimation of the bromine absorption of wines, the values obtained being of certain use in determining the quality of a wine (see below).Ten C.C. of the wine are treated with 20 C.C. of water, 20 C.C. of & bromate solution (0.57 grm. of potassium bromate and 2 grms. of potassium bromide per litre), and 7 C.C.of 20 per cent. hydrochloric acid; in the case of red wines, 40 C.C. of the bromate solution are added, and the water is omitted. After the lapse of exactly five minutes, potassium iodide solution is added, and the mixture is titrated with TT thiosulphate solution. A blank estimation, using 30 C.C. of water, is made at the same time. The diiference between the two estimations is termed ‘‘ bromine absorption A.” Another portion of 50 C.C.of the wine is then mixed with 1 grm. of calcium carbonate, and distilled until about one-third of the liquid remains in the flask : after cooling, this residual liquid is diluted to 25 c.c., and 10 C.C. of the turbid solution are treated with 4 C.C. of a 19 per cent. lead acetate solution and 6 C.C. of water.The mixture is shaken, filtered, and the bromine absorption is estimated on 10 C.C. of the filtrate ; the result gives “ bromine absorption B.” Another 10 C.C. of the turbid solution are treatedFOOD AND DRUGS ANALYSIS 267 with 4 C.C. of basic lead acetate solution (sp. gr. 1.24) and 6 C.C. of water, filtered, and the bromine absorption estimated on 10 C.C. of the filtrate ; this gives “ bromine absorption C.” The difference between ( 6 A ” and ‘‘ B” corresponds with the quantity of tannin and colouring matters present in the wine ; also, iu certain cases, to the gallic acid and various chlorophyll derivatives.The difference between “ B ” and “ C” is due to substances which are also probably chlorophyll derivatives or complexes, whilst the residual ( ( bromine absorption (‘C” is caused by such wine constituents as yield a red coloration with vanillin and hydrochloric acid.The following results obtained with red and white wines are recorded : <(A,” erg." ( I C.” White wines (37 samples) ... . . . 4.53-11.47 2.62-4‘85 2’04-3.55 Red wines (10 samples) ... . . . 5’52-22.75 3.26-7.22 2.00-3*33 The values “ A ” and ‘( B ” are decreased when the wines are clarified or ‘( sulphured,” but “ C” is not affected ; whilst all the values are diminished when the wine contains added water. w.P. s. are not defined in the paper from which this abstract was taken.-W. P. S. Note by Abstractor.-The terms in which the bromine absorptions are expressed Estimation of Quercitin in Wine. T. von Fellenberg. (Mitt. Leben- smitteluntersach. Hyg., 1913, 4, 1-14 ; through Chem.Zentralbl., 1913, I., 1366-1367.) -By heating under pressure 1 kilo. of vine-shoots with 3 litres of water for thirty minutes to a temperature of 130° C., filtering the solution, and extracting the filtrate with ether, 0.17 grm. of colouring matter was obtained in the form of crystals; the latter were needle-shaped, melted at about 250” C., and gave all the reactions of quercitin (pentahydroxyflavonol).The yellow colouring matter present in wines may be extracted by means of ether, then fixed on mordanted wool, and its quantity estimated by comparison with the colour obtained under similar conditions with known amounts of the substance. I t appears that only traces of quercitin occur as such in wine, and that it is quercitrin (the methyl-pentoside of quercitin) which is present.Red wine contains a brownish-red colouring matter, which may be removed by fixing it on unmordanted wool before the quercitin is dyed on to mordanted wool. Fresh grape-juice does not usually contain the yellow colouring matter, but the latter is formed gradually during fermentation, especially in the case of the juice from grape-skins. White wines do not contain more than 4 mgrms.per litre of the substance, whilst red wines contain much more, as do also both red and white wines prepared from grape-skins, etc.; the presence of more than 8 mgrms. of yellow colouring matter in a litre of white wine indicates that the wine has been prepared from grape-skin juice. This yellow colouring matter of wines probably consists of quercitrin, quercitin, and a, decomposition product of a chlorophyll complex.w. I?. s. Composition of Vermouth Wines. A. Behre and K. Frerichs. (Zeitsch. Untersuch. Nahr. Genussm., 1913, 25, 429-444.)-Results of analyses of twelve samples of vermouth wine (a stimulating liquor usually consisting of wine, to which268 ABSTRACTS OF CHEMICAL PAPERS have been added absinthe, angelica, and other aromatic substances) are recorded. The minimum and maximum figures obtained were-Alcohol, 14.45 and 16.93 per cent. by volume; total solids, 9.88 and 16.75 per cent.; ash, 0.081 and 0.258 per cent. ; total acid (as tartaric acid) 0.281 and 0.473 per cent. ; tartaric acid, 0.072 and 0.191 per cent.; glycerol, 0.244 and 0.669 per cent.; invert sugar, 1.77 and 14-16 per cent. ; sucrose, 0 and 11.26 per cent, It appears to be exceedingly difficult to ascertain by analysis how much wine is actually present in these beverages, especially in cases where the proportion of added liquid is not excessive. w. P. s.

ISSN:0003-2654    

DOI:10.1039/AN9133800264

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

268 ABSTRACTS OF CHEMICAL PAPERS BACTERIOLOGICAL, PHYSIOLOGICAL, ETC. Distribution of Ante-Mortem Administered Arsenic in the Human Body. J. B. Ekeley. (J. Amer. Chem. Xoc., 1913, 35, 483-485.)-The following results were obtained from an analysis of the organs of a person known to have died from the effects of criminally administered arsenic. The body when analysed was in an advanced state of moist decomposition; the stomach, liver, kidneys, and heart had been removed from the body previous to burial, and traces of arsenic wera found in the soil adjacent to the thin casket containing the corpse.Johnson and Chittenden’s modification of Gautier’s metkiod, in which only volatile reagents are employed, was found very satisfactory. The arsenic actually found, which was only a portion of the total amount present in the body, was as follows : 60 grms. stomach contents, 0.0177 grm.arsenic ; 72 grms. stomach walls, 0.0124 grm. arsenic; 821 grms. liver, 0.0598 grm. arsenic; 763 grms. intestines, 0.218 grm. arsenic ; 193 grms. kidneyg, 0.0355 grm. arsenic; 127 grms. heart, 0-0012 grm. arsenic; 932 grms. brain, 0.009 grm. arsenic; 18 grms. spinal cord, trace of arsenic ; 100 grms.thigh, 0.0003 grm. arsenic ; 82 grms. toes, 0.0002 grm. arsenic. If the figures are calculated so as to show arsenic per 100 grms. of tissue used, by far the greater part is found to be equally divided between the kidneys and stomach walls. These results are at variance with those of Scolosuboff (BUZZ. Soc. Chirn., (a), 24, 124), who found most arsenic in the spinal marrow, then in the brain, next the liver, and least in the muscles; but they agree in general with those of Ludwig (Med.Jahrbuch, 1880) and with Chittenden ( J . Amer. Chem. SOC., 1883, 5,8). The contention of the last author that ‘( the finding of arsenic in the brain is an indication amounting to almost positive proof that the poison was not post-mortem ” would seem to be confirmed. H.F. E. H. Total, 0.1498 grm. arsenic. Action of ‘‘ Activated ” Aluminium on Alkaloid Extracts and its Use in Toxicology. E. Kohn-Abrest, Rochas, and Rivera-Maltes. (Ann. Fulsq., 1913, 6, 131-138.)-Aluminium which has been treated with mercuric chloride may be employed for clarifying alcoholic extracts of alkaloids, such as are obtained in the separation of alkaloids from decomposing substances in the course of toxico- logical analysis.The aluminium is rendered active by immersing it for three minutes in a 1 per cent. solution df mercuric chloride, and then washing it thoroughly. The aluminium may be used in the form of turnings, but small pieces of foil will beBACTERIOLOGICAL, PHYSIOLOGICAL, ETC. 269 found more convenient.Preliminary experiments made by the authors show that when alcoholic solutions of alkaloids areitreated with the (‘ activated ” aluminium, the alkaloid is not precipitated except in the case of strychnine, quinine, and cocaine, which are precipitated partially, whilst nicotine is removed completely from its solution. I n testing viscera, etc., for the presence of alkaloids, the usual extraction is made with alcohol in the presence of tartaric acid, the alcoholic extract is filtered and treated with ‘6 activated ” aluminium for twelve hours.More acid must be added if the mixture becomes alkaline during this period. The solution is then filtered, and the alkaloids are isolated from the filtrate in the usual way, and identified. w. P. s. Estimation of Colloids in Effluents.P. Rohland. (Zeitsch. Chem. Ind. Kolloide, 1913, 12, 45-46 ; through Chem. Zentralbl., 1913, I., 1539-1540.)-The most injurious substances in trade effluents, especially from sulphite cellulose works, are of a colloidal nature. For their estimation 50 to 100 C.C. of the effluent, which has been freed from suspended particles, are treated with 1 C.C. of a 1 per cent.solution of aniline blue, and evaporated to a syrup on the water-bath. The residue is diluted with water and filtered through a tared filter, the insoluble matter washed with water and weighed, and the amount of absorbed dyestuff deducted from the weight. The latter value is found by a colorimetric estimation of the unabsorbed dyestuff. C. A. M. Attempt to Estimate the Vitamine Fraction in Milk.C. Funk. (Biochem. J., 1913, 7, 211-213.)-The whole vitamine fraction (nitrogen precipitated by phosphotungstic acid in the alcoholic extract of the dried milk) from 1 litre of milk amounts to 1 to 2.5 mgrms. nitrogen. This would correspond to 10 to 30 mgrms. vitamine (C17H,,07N2). After the elimination of the vitamine fraction the residual nitrogen amounts to 20 to 50 mgrms.per litre of milk. This residual nitrogen represents, in all probability, allantoin; if so, then 1 litre of milk contains 0.06 to 0.15 grm. allantoin. Milk, after removal of fat by centrifuging, loses about 50 per cent. of vitamine and allantoin. The milk used in the experiments was obtained from London dairies, and the present investigation shows the ordinary chemical methods for estimating vitamines hardly suffice, and attention must be directed to colorimetric methods.The behaviour of allantoin when boiled with triketohydrindene hydrate (Ruhemann, J. Chem. SOC., 1910, 97, 2025) is of interest. Abderhalden and Schmidt (Zeitsch. physiol. Chem., 1911, 72, 87) state that when these substances are boiled together for one minute no coloration results. The authors confirm this, but state that continued boiling (two to three minutes) causes a violet coloration, which might be used as a, test for allantoin. Allantoin, like vitamine, is destroyed by boiling.H. F. E. H. Estimation of Urea in Urine by Sodium Hypobromite. M. Krogh. (Zeitsch. Physiol. Chem., 1913, 84, 379-408.)-The reaction between sodium hypo- bromite and urea does not proceed completely according to the generally accepted equation : CO(NHJ, + 3NaOBr + 2NaOH = Na2C0,+ 3NaBr + 3H20 + N,.270 ABSTRACTS OF CHEMICAL PAPERS Part of the nitrogen is converted into oxygen compounds which were not identified, and part of the carbon is only oxidised to carbon monoxide.The greater the ratio of bromine to sodium hydroxide, the more far-reaching is the oxidation. If glucose is present the method is inaccurate, as the volume of evolved gas is increased, owing to the nitrogen being completely liberated and a greater proportion of carbon being converted into monoxide.On decomposing ammonium chloride with sodium hypo- bromite, only part of the nitrogen is liberated, a part being converted into oxygen compounds. The concentration of the urea or ammonia does not affect the reaction, provided an excess of sodium hypobromite is present, in which case the liquid remains yellow after the reaction.Changes of temperature up to 5 O C. do not affect the reaction. Certain constituents of urine other than urea are also decomposed by sodium hypo- bromite, with evolution of nitrogen. This source of error can be diminished, though not entirely eliminated, by precipitating the urine with phosphotungstic acid and treating the neutralised filtrate with hypobromite.For the determination of urea in urine the following method is recommended: 5 C.C. urine are titrated with a 10 per cent. phosphotungstic acid solution in 4 sulphuric acid to determine the amount required for complete precipitation.The amount of urine which, on addition of the requisite volume of phosphotungstic acid, will yield 10 c.c., is measured into a 10 C.C. graduated flask, phosphotungstic acid added, filtered, the filtrate neutralised with 30 per cent. sodium hydroxide and decomposed with a solution of 1 C.C. bromine in 24.5 C.C. 30 per cent. sodium hydroxide 1.00 96.5 + 70 C.C. water. The quantity of gas evolved x- = urea nitrogen.If the hypo- bromite be prepared by dissolving 1 C.C. bromine in 196 C.C. 30 per cent. sodium hydroxide, the evolved gas corresponds exactly with urea nitrogen. The results obtained by this method are lower than those obtained by decomposing the phospho- tungstic filtrate under pressure according to Folin: this is probably due to the presence of such substances as oxyprotein acids, which are not removed by phospbo- tungstic acid, nor decomposed by hypobromite, but are decomposed at high tem- peratures.The results obtained by the hypobromite method are more likely to be too high than too low, since phosphotungstic acid does not remove all substances decomposable by hypobromite. E. W. Estimation of Chlorides in Urine and Blood.K. 0. Larsson. (Biochem. Zeitsch., 1913, 49, 479-486.) - Bang has found that all substances occurring in urine which react with silver nitrate other than chlorides can be removed by means of blood charcoal, and that the chlorides in the filtrate can be accurately determined by Mohr’s method. The author has worked out the details of the method and also its application to blood.For urine 20 C.C. are shaken with 1 grm. blood charcoal for ten minutes, filtered, and 10 C.C. of the filtrate titrated with If the sp. gr. of the urine is higher than 1.025 it must be diluted with water. I t is necessary that the urine be acid. For blood, 5 to 10 grms. are treated with about 50 C.C. of a boiling 2 per cent. magnesium sulphate solution, a few drops of 40 per cent, acetic acid added, and the mixture silver nitrate, using potassium chromate as indicator.The quantity of charcoal may not exceed 2.5 grms. per 20 C.C.ORGANIC ANALYSIS 271 heated on a water-bath until coagulation is complete. After cooling, the mixture is made up to 100 c.c., shaken, and after two hours poured on to 3 grms. blood charcoal, shaken, and after a couple of minutes filtered, and the chlorine in 50 C.C. filtrate estimated by Mohr’s method. The method gives exact results if 5 to 10 grms. blood are used, but if smaller quantities, 1 to 3 grms., are used, the results are too high, owing to the solutions being too dilute. This source of error can be eliminated by concentrating the liquid before titration. The coagulation of the albumin is necessary. If it be omitted, the results are 20 to 30 per cent. too low, which possibly may indicate that the salt occurs in combination with the albumin. E. W.

ISSN:0003-2654    

DOI:10.1039/AN9133800268

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

ORGANIC ANALYSIS 271 ORGANIC ANALYSIS. Estimation of Carbon by Combustion with Moist Oxygen. S. Hilpert. (Ber., 1913, 46, 949-952.)-1t is known that the presence of water-vapour materially assists the combustion of difficultly combustible substances. The author has applied this principle in the combustion analysis of certain carbon compounds. Magnesium and manganese carbides containing a high percentage of carbon can only be burnt very incompletely in dry oxygen, but are readily dealt with if the current of oxygen be first passed through a washing-bottle containing water.Naturally, an efficient calcium chloride tube must be provided in front of the potash bulbs, in order to remove the moisture after combustion. Another important application of this prih- ciple is for the determination of carbon in steels.I n the ordinary way this can only be burnt at the very high temperature of 1,200' C. in the electric furnace. With the use of moist oxygen, however, a temperature of 950' C. is amply sufficient in most cases to produce a complete combustion of the carbon. Only in the case of an 80 per cent. ferromanganese with 6-16 per cent. of carbon was a temperature as high as 1,050' C., with a prolonged heating, required for the analysis. The use of moist oxygen is also advantageous in certain cases in the analysis of organic substances which form hard masses of carbon or of explosive compounds. Water-vapour is very useful in the combustion of aluminium-phenyl and of diazo compounds.Nitrogenous compounds, such as hippuric acid, glycocoll, alanine, and carbazol, may be burnt completely and with accuracy without the necessity of mixing copper oxide with the substance in the boat.If it is required to determine the hydrogen in the same portion of the substance, arrangements must be made to conduct the first part of the combustion in the ordinary way, then change the calcium chloride tube, and finish up with moist oxygen.It is desirable to insert a straight tube about 5 cm. long, filled with phosphorus pentoxide, between the calcium chloride tube and the potash- bulbs, in order to absorb the last traces of moisture. Finally, the moist oxygen gives very good results when employed for the combustion analysis of substances such as phenyl-phosphoric acid, in which the carbon becomes coated with fusible ash.In the ordinary way only a very careful mixing of the substance with much copper oxide will give accurate results, but with moist oxygen the substance is readily burnt, leaving a residue of perfectly white metaphosphoric acid. J. F. B.272 ABSTRACTS OF CHEMICAL PAPERS Experiments on the Direct Estimation of Caoutchouc by Titration with Bromine. F.Kirchhof. (Gummi. Zeit., 1913, 27, 9 ; through Chem. ZentrulbZ., 1913, I., 1467-1468.)-Five grms. of extracted raw Para rubber are dissolved in a stoppered bottle in about 200 C.C. of carbon tetrachloride, and treated with an excess of After a time of action varying from 4 to 12 hours, with repeated shaking, a concentrated aqueous solution of potassium iodide is added, and after complete reduction of the excess of bromine, the separated iodine is titrated with A- thiosulphate, using starch as an indicator. It was found that the reaction afforded no sharp end-point for the titration, since the decolourised solution always turned blue again on standing.Further, it was found that the quantity of bromine absorbed by the caoutchouc depended largely on the time allowed for its action.The observed values ranged from 94.4 to 117 per cent. calculated as CloHl,. The quantity of hydrobromic acid formed varied according to the purity and dryness of the reagents, the temperature, concentration, time, and other conditions of the action. No better results were obtained with vulcanised rubber. J. F. B. bromine solution (in carbon tetrachloride).Method for the Estimation of the Nitrogenous Constituents of Raw Caoutchouc. A. Tschirch and W. Schmitz. (Gammi. Zeit., 1912,26, 2079-2081 ; through Chem. Zentrulbl., 1913, I., 1467.)-From raw Para, rubber the authors obtained 2.9 to 3-1 per cent. ofa residue insoluble in pentachlorethane, which contained 11 to 12.6 per cent. of nitrogen. The raw rubber itself contained 0.3571 per cent.of nitrogen, of which 0.345 per cent. was found in the residue from the pentachlorethane extraction. After deducting the ash, the nitrogen in the residue amounted to 14 to 15.6 per cent., which is very nearly that contained in a pure protein. Neverthe- less this residue does not show the well-defined reactions characteristic of proteins, so that the authors are inclined to think that the nitrogenous constituents of crude rubber may not be true proteins. J.F. B. Estimation of the Viscosity of Rubber Solutions. J. G. Fol. (Chem. Weekblud., 1913, 10, 154-174 ; through Chenz. Zerztralbl., 1913, I., 1369.)-The author reviews various methods which have been proposed, particularly that of Schidrowitz and Goldsbrough (ANALYST, 1909, 34, 112), and gives the results of his own expsri- ence as to the estimation of the viscosity of rubber solutions. H e comes to the conclusion that Ostwald's capillary viscometer is the most suitable apparatus for the purpose, but the diameter of the capillary must not be too small.The viscosity is affected if the solution is shaken violently during its preparation, and there is a considerable difference in the relative viscosities of solutions of equ_al concentration in which the rubber is dissolved completely and incompletely, respectively.The viscosity also varies with the age of the solution, and the absolute viscosity, but not the relative viscosity, of Hevea rubber alters with the temperature. In the case of the more concentrated solutions, the time of efflux diminishes while successive determinations are being made.The following method is recommended : The rubber is cut up into small pieces, dried in vucuo over sulphuric acid, and portions of 0.025, 0.5, and 1.0 grm. are weighed into brown glass bottles. After 100 C.C. of pureORGANIC ANALYSIS 273 benzene have been added to each bottle, the contents are shaken twice a day for three days, then filtered through glass-wool, allowed to stand for another day, and the viscosity then estimated at 30' C.in an Ostwald's capillary viscometer of standard dimensions ; three readings are taken for each solution, and the concentra- tion of the same is found by evaporating portions and weighing the residues. A viscosity curve is now plotted with a standard scale (1 C.C. on the x-axis = 2 viscosity units; 1 C.C. on the y-axis = 0.1 per cent.concentration). The viscosity number, which serves as a measure of comparison between different rubbers, is represented by the area (in sq. cm.) between the co-ordinates, the viscosity curve, and the ordinate corresponding with 1 per cent. concentration. w. P. s. Analysis of Mixtures of the Lower Fatty Acids. K. Langheld and A.Zeileis. (Bey., 1913, 46, 1171-1177.)-An accurate method for the separation of the lower fatty acids is required, particularly in the investigation of the group of monoamino-acids formed by the hydrolysis of proteins. The authors have devised a method of fractional moist combustion by means of chromic acid, which they have developed for the analysis of mixtures containing four typical acids of the series- viz., acetic, isobutyric, isovaleric, and methyl ethyl acetic acids.Using chromic and sulphuric acid mixture, it is found that the oxidation of the acids may be split up into three stages, corresponding to the temperatures of 65', looo, and 170' C., and that the branch-chains of the iso-acids are oxidised at the lower temperatures. The following table shows the number of molecules of carbon dioxide split off from the type acids at various temperatures : At 6 5 O C.At 100" C. At 1700 C. 2 Acetic acid ... ... ... Isobutyric acid ... ... ... 1 1 2 Isovaleric acid ... ... ... 2 1 2 Methyl ethyl acetic acid ... 1 1 3 - - The determination of the carbon dioxide produced at these three stages and the total acidity of the mixture give four data, from which the percentage of each of the four acids in the mixture may be calculated.Thus, let a, b, and c, represent the weights of carbon dioxide produced at the respective temperatures above, and let t represent the number of C.C. of fc alkali required to neutralise the total acidity of the mixture, then : Isobutyric acid=-. 2 b - a - c + 2 . 44*t ) grm. 4848 ( 10000 Isovaleric acid = 120 , (a - b ) grm.44 Methyl ethyl acetic acid=- c - 2 - 44.t ) grm. Io2 44 ( 10000 The method is carried out in the following manner : The oxidation is performed in the apparatus devised by Messinger for moist combustions. The acid mixture is first274 ABSTRACTS OF CHEMICAL PAPERS accurately titrltted with & alkali, and a quantity of the neutralised solution, equiva- lent to 60 to 70 C.C.of & alkali, is taken for the combustion. The liquid is concen- trated to about 25 C.C. and transferred to the combustion flask, To this are added, whilst cooling in a freezing mixture, 100 C.C. of concentrated sulphuric acid, and the liquid is allowed gradually to assume the temperature of the atmosphere. Meanwhile, air freed from carbon dioxide is passed through the apparatus, and when the latter is free from carbon dioxide the absorption flask is connected up.A solution of 7 grms. of chromic acid in 25 C.O. of water, mixed with 100 C.C. of concentrated sulphuric acid, is then admitted through a dropping funnel, and the reaction flask is placed in a water-bath at 65" C., which is kept at this temperature until the evolution of carbon dioxide at this stage is finished ; this requires about thirty minutes.The apparatus is then cleared out by a current of air free from carbon dioxide, and the absorption vessel is weighed. The same operations are repeated at the other two stages of the combustion at temperatures of 100" and 170" C. respectively. It is to be noted that, for the complete oxidation of acetic acid and propionic acid, about three hours are generally required. The results are sufficiently good to afford very useful indications not readily obtainable by older methods, and the process possesses the advantage that only small quantities of the material are required for the analysis.J. E. B. Detection of Volatile Fatty Acids. H. Agulhon. (BUZZ. Soc.Chim., 1913, 13, 404-408.)-When the aqueous solution of the sodium salt of a fatty acid is treated with copper sulphate and then shaken with certain organic solvents, a whole or part of the copper salt of the fatty acid may be extracted, the proportion removed being a function of the molecular weight of the fatty acid. For example, none of the usual solvents extract the copper compound in tho case of sodium formate or acetate.With the propionate there is a slight blue coloration on extracting the solution with ethyl acetate, but no coloration with ether, chloroform, or benzene. Sodium butyrate, however, gives an intense blue coloration with ethyl acetate, ether, chloroform, and amyl alcohol, but not with benzene or toluene if the proportion of the sodium salt did not exceed 1 or 2 per cent. ; whilst in the case of sodium valerate or caproate all the solvents remove the whole of the copper salt.Similar differences are to be observed in the behaviour of the iron salts of the different fatty acids. In applying the method, the liquid, which must not contain more than 2 per cent. of fatty acid, is exactly neutralised with a standard solution of sodium hydroxide, with phenolphthalein as indicator, and 2 C.C.of the solution are taken for each test. About 1 C.C. of ether is added, and then, drop by drop, with continual shaking, a 2 per cent. solution of copper sulphate. A blue coloration of the ether indicates a butyrate, valerate, or the salt of a higher fatty acid. The test is now repeated with 1 C.C. of benzene, andif the latter remains colourless, butyric acid is probably present, whilst a blue coloration denotes the presence of valeric acid or a higher fatty acid.If in the first test the ether remained colourless, the test is repeated with ethyl acetate and a dilute (less than 5 per cent.) solution of ferric chloride, when a yellow coloration of the solvent indicates propionic acid, whilst with acetic acid it will still be colourless.C. A. M.ORGANIC ANALYSIS 275 Detection of Formic and Acetic Acids. L. Bonnes. (Bull. Sci. Pharmacol., 1913, 20, 99-101 ; through Chem. Zentralbl., 1913, I., 1364.)-Formic and acetic acids may be detected in such substances as glycerol by distilling about 4 C.C. of the sample with an excess of sulphuric acid, neutralising the distillate by the addition of an exceas of calcium carbonate, and evaporating the mixture to dryness.The residue is then subjected to dry distillation, and the distillate is collected in a small quantity of water. A portion of this solution is tested for the presence of aldehyde with Ley's rosaniline bisulphite solution, and another portion for acetone by the sodium nitro- prusside test.If Ley's solution gives a positive reaction, formic acid is present in the sample, although a negative reaction may result with the nitroprusside test; the presence of acetic acid is denoted by a positive reaction with nitroprusside and a negative reaction with Ley's solution, Positive reactions with both tests indicate the presence of the two acids together. w. P. s.Pine-nut Oil. M. Adams and A. Holmes. (J. Ind. and Eizg. Chem. 1913, 5, 285-287.)-The kernels of the nuts of the pine-nut tree, Pznus monophylla, which grows on the eastern side of the Sierra Nevada Mountains, yield an oil resembling olive oil, with a bland aromatic taste. The freshly extracted oil is light yellow, but the colour gradually fades. A sample of the nuts yielded 12.4 per cent.of oil (16.2 per cent. on the kernels). This oil melted at -15' C. and distilled without decomposition at 305" C. under a reduced pressure of 60 mm. It gave the following values : Refractive index at 15' C., 1.4733 ; at 40' C., 1.4543 ; saponification value, 189.31 ; iodine value (Hiibl), 107.9 ; iodine value of L L liquid " fatty acids, 117.1 ; and glycerol, 9.2 per cent.The fatty acids consisted of oleic acid, with small amounts of stearic (1 per cent.), palmitic, lauric, and linolic acid. No linolenic acid was found, but there was apparently some other unsaturated fatty acid, which was not identified. No phytosterol could be detected. C. A. M. New Method of Estimating Glycerol. S. H. Bertram. (Chem. TVeelcblad., 1913, 10, 237-239 ; through Chem.Zentralbl., 1913, I., 1543.)-The solution containing the glycerol is treated with excess of potassium hydroxide solution and 10 per cent. copper sulphate solution until a permanent precipitate is obtained. The liquid is then acidified with acetic acid, an excess of potassium iodide solution added, and the separated iodine titrated with Tv thiosulphate solution. Each molecule of sodium thiosulphate corresponds to two molecules of glycerol.(Cf. Muter, ANALYST, 1881, 6, 41.) C. A. M. Heat Test. I.-Guncotton. 11.-Nitroglycerine and Cordite. 111.- Theoretical. A. C. Egerton. (J. Soc. Chem. Ind., 1913, 32, 331.)-Several weak points in the Abel heat test are pointed out, and a new test is described, in which the paper is replaced by a drop of a-naphthylamine and sulphanilic acid in acetic acid, or of dimethylaniline in acetic acid.The drop, of definite surface, is contained in a small cell placed in a compartment attached to the ordinary heat-test tube. With the paper indicator, a guncotton continuously tested gives first (1) a fall and then (2) a rise of heat test. Differing from Smart and Robertson (ANALYST, 1910, 35,276 ABSTRACTS OF CHEMICAL PAPERS 171) who regard this as due to (1) nitric peroxide retained by the heated guncotton, and (2) to a drying of the test paper, the author attributes it to the formation of nitric acid from the peroxide and the moisture in the guncotton ; at first (1) there is sufficient moisture, and the test falls ; but (2) on continued heating of the guncotton it loses moisture, less nitric acid is formed, and the test rises, The fall is less marked with the drop-test.Experiments on the variation of moisture during the test and on the effect of evacuation are given in support of this theory. A continu- ous test is stated to give much more valuable results than a single one. With nitro- glycerine and cordite the presence of several test-papers or drops produces little alteration in the test, although the amount of nitric peroxide absorbed by them is not negligible ; the setting up of an equilibrium between the partial pressure of the nitric peroxide and the dissociation pressure of the explosive at the test temperature is assumed to account for this.The equilibrium pressure is stated to be a more certain measure of stability than the Abel test, and it is maintained that the curve given by a continuous test affords a satisfactory indication of the tendency of an explosive to decompose.0. E. M. Analysis of Mixtures of Hydrogen with Methane, Ethane, Propane, and Isobutane. P. Lebeau and A. Damiens. (BUZZ. SOC. Chim., 1913, 13, 366-372.)--The paper describes the authors' work with mixtures of hydrogen, methane, ethane, and propane (ANALYST, 1913, 121), and, in addition, shows how, by a modification of their method, mixtures containing these gases together with iso- propane may be analysed.The hydrogen and methane are first separated in the manner already described (Zoc. cit.), and separately estimated in any convenient manner. The residual liquid-consisting of ethane, propane, and isobutane-is then allowed to attain a temperature of - 127" C.by substituting for the bath of liquid air a bath of light petroleuni cooled to -127" C., and kept at, or about, that tmperature by being enclosed in a Dewar vessel, into which liquid air is introduced drop by drop, so that it always stands at a marked level, The temperature of the petroleum bath is determined by the vertical height between its base and the surface of the liquid air, and is measured by means of a calibrated thermo-couple.The temperature may rise to - 123.5" C. without detectable error. With the temperature between - 127' and - 1 2 3 ~ 5 ~ C., the pump is set in operation and the gas, which contains all the ethane and some propane, is removed and analysed eudiometrically. The residual liquid, which contains all the isobutane and the rest of the propane, is then allowed to a~sume the gaseous state and analysed similarly to the other binary mix t ures.G. C. J. Colour Reactions of Certain Indole Derivatives and their Significance with regard to the Glyoxylic Reaction. A. Homer. (Biochem. J., 1913, 7, 116-126.)-The formation of coloured condensation products from indole derivatives and certain aldebydes necessitates the use of condensing agents. The reaction takes place between the - NH group of the indole nucleus and the - CHO group of the aldehyde.The evidence adduced in this paper is favourable to the Adamkiewicz reaction being primarily a formaldehyde reaction. The formaldehyde condensationORGANIC ANALYSIS 277 product of tryptophane gives the colour test directly with concentrated sulphuric acid without the addition of an oxidising agent or of glyoxylic acid or of formaldehyde.A trace of formaldehyde without the addition of an oxidising agent gives the colour reaction with indole derivatives. When forrnaldohyde is used in testing for indole derivatives, the purple colour is often masked on account of further reactions taking place between the indole substance and the trioxymethylene formed from formaldehyde by the condensing power of the concentrated acid.In Hopkins and Cole’s modification of the test, glyoxylic acid reacts by virtue of its decomposition into formaldehyde, and for one of the two following reasons is the best reagent to use for the reaction : (a) The formaldehyde is liberated in such small quantity that the formation of trioxymethylene does not take place to any appreciable extent ; (b) the formaldehyde so liberated is in the nascent state, and therefore more reactive.In the case of tryptophane the colour test is often performed with formaldehyde in the presence of an oxidising agent. The part played by the latter is primarily to produce some oxidation product of tryptophane-e.g., indolealdehyde.The reaction under tbese conditions is of a complex nature, as besides the effects described above, it involves the formation of coloured substances by the action of the concentrated acid on ( a ) indolealdehyde and tryptophane; ( b ) indolealdehyde in the presence of an oxidising agent ; (c) indolealdehyde and tryptophane in the presence of an oxidisiag agent ; ( d ) indolealdehyde and tryptophane in the presence of formaldehyde.H. F. E. H. Application of the Dimethyl Sulphate Test for Determining Small Amounts of Petroleum or Asphalt Products in Tars. C. S. Reeve and R. H. Lewis. (J. Ind. and Eng. Chem., 1913,5,293-295.)-Valenta stated (ANALYST, 1906, 31, 202) that aromatic hydrocarbons were miscible in all proportions with dimethyl sulphate, whereas open-chain hydrocarbons were entirely insoluble in that reagent. Chapin has shown (U.S.Dept. Agric., CircuZar 167) that open-chain hydrocarbons show varying degrees of solubility in dimethyl sulphate, and that when a mixture of aromatic and open-chain hydrocarbons is shaken with the reagent, a varying, though usually small, amount of the aromatic hydrocarbons remains dis- solved in the upper (paraffin) layer.Nevertheless, he pronounced the test of value, and Sommer has attempted to apply it to the determination of asphalt in tar-asphalt paving mixture. Sommer (J. Ind. and Eng. Chem., 1910, 2, 181) showed that the distillates from native bitumens (largely polymethylenes) were insoluble in dimethyl sulphate, and he proposed to distil the tar-paving mixture to coke, and to test 4 C.C.of the distillate by shaking with 6 C.C. of the reagent ir, a graduated cylinder. In the distillate from a mixture containing 80 per cent. asphalt, he found 60 per cent. insoluble, whilst a 20 per cent. mixture (the poorest he tried) gave 15 per cent insoluble. The present authors failed to get any evidence of the presence of asphalt when applying Sommer’s method to 10 per cent.paving mixtures, and consider that this is not surprising, as a distillate obtained in Sommer’s manner might be expected to contain a much higher proportion of aromatic hydrocarbons than the original mixture. As some current specifications provide for smaller proportions than 10 per cent., and Bince the control of deliveries to such specifications is extremely difficult278 ABSTRACTS OF CHEMICAL PAPERS in the present state of our knowledge, the test was further investigated and modified as follows : The sample (100 c.c.) is distilled, and the fractions distilling below 315', between 315" to 350', and between 350' to 375' C., are collected separately.The two higher fractions are then measured and submitted to Sommer's test of shaking 4 C.C.with 6 C.C. of dimethyl sulphate. Experiments with asphalts of various origin show that the fraction distilling below 315" C. is unlikely to contain a detectable trace of asphalt, even when the product is a high-grade one. The modified method serves to detect asphalt in mixtures containing as little as 3 per cent. I t can only be used for the approximate estimation of the amount of asphalt if the grade and character of the asphalt is known, for different asphalts may yield anything from 30 t o 50 per cent.of distillate below 375' C. The proportion of the distillate insoluble in dimethyl sulphate is seldom less than 80 per cent., and never more than 90 per cent. Provided the character of the asphalt is known, the figures in the paper show that its amount might be estimated with an error not greater than ~t: 20 per cent., or that failure to find as much as 8 per cent.where 10 per cent, was specified would be tolerably strong evidence that the sample was not up to specifica- tion. Until a more exact method is forthcoming, the authors think that their test, in conjunction with a volatilisation test to determine the character of the material which has been fluxed with a tar, will prove useful.Failing more exact methods, they suggest that the chemist's duty in seeing that specifications are complied with would be more easily fulfilled if the grade and character of the asphalt product and tar to be used were specified as well as their relative proportions.G. C. J. Hydrolysis of Organic Phosphorus Compounds by Dilute Acid and Dilute Alkali. R. H. A. Plimmer. (Biochem. J., 1913, 7, 72-80.)-Plimmer and Scott (J. Chem. fi'oc., 1908, 93, 1699) have shown that the phosphoproteins may be distinguished from other organic phosphorus compounds owing to the fact that the phosphorus of the former maybe completely separated as inorganic phosphate in twenty- four hours by treatment with 1 per cent.caustic soda at 37' C. Under these conditions soda does not decompose the other natural organic phosphorus compounds. The action of dilute acid and alkali from normal to twice normal strength has been studied upon glycerophosphoric acid, phytic acid, ethyl phosphoric acid, hexosephosphoric acid, nucleic acid, hydroxymethyl-phosphoric acid, and phosphoprotein. The analyses of phosphoric acid were made by precipitation with ammonium magnesium citrate and conversion into magnesium pyrophosphate.Total phosphoric acid in the solution was estimated in a separate sample by Neumann's method (J. Physiol., 1906, 33, 439). Inorganic P,O, not being precipitated in the presence of phytic acid, was estimated by precipitation with ammonium molybdate. Nitric acid was used for the acid hydrolyses.Glycerophosphoric acid is slowly hydrolysed by dilute acid, but is not attacked after eighty-six days by alkali. Ethyl phosphoric acid is also stable to alkali, but is hydrolysed by acid more slowly than glycerophosphoric acid. Phytic acid is the most stable of all the organic phosphorus compounds, complete hydrolysis not being effected even by f nitric acid in seventeen days at 75O C.alkali has no action upon phytic acid after thirty-two days at 75' C. Hexosephosphoric acid is easily hydrolysed by both acid and alkali. Nucleic acid is completely hydrolysedORGANIC ANALYSIS 279 I I I by caustic soda in thirty-two days. Hydroxymethyi-phosphoric acid is not hydrolysed by dilute acid at 75" C.Stability to alkali is a property of the esters of phosphoric acid, while hexosephosphoric acid and phosphoprotein are so different in their behaviour to alkali that some difference in their constitution from that of the esters must exist. Nucleic acid, since it is hydrolysed by both acid and alkali, seems to occupy an intermediate position between the stable esters and the unstable hexosephosphoric acid.H. F. E. H. hydrochloric acid in eight days at 75" C., and about half hydrolysed by Density of Solution. Investigation of Phytin. R. H. A. Plimmer and H. J. Page. (Biochem. J., 1913, 7, 157-173.)-Phytin is generally regarded as the calcium magnesium salt of inositol phosphoric acid (phytic acid). Commercial phytin, prepared from oil-cake by extraction with 0.5 to 1 per cent.of hydrochloric acid, was used almost entirely, but for comparison neutral calcium phytate was prepared from bran. The large amount of experimental work described may be summarised as follows: (1) Inorganic phosphates in phytin can be readily estimated by precipitation with ammonium rnolybdate in seminormal nitric acid at room temperature.(2) The calcium content of phytin cannot be estimated by precipitation as oxalate, but is easily ascertained by precipitation as calcium sulphate. The magnesium content of phytin can then be estimated as magnesium pyrophosphate. (3) There is great difficulty in removing the calcium from phytin in the preparation of phytic acid. (4) The yield of inositol, obtained in the hydrolysis of:phytic acid by acids, is not quantitative; it is probable that there is another organic constituent present in phytin.€I. F. E. H. Calculated Density of Protein. Density and Solution Volume of some Proteins. H. Chick and C. J. Martin. (Biochem. J., 1913, 7, 92-96.)-The authors have previously pointed out (Zeeitsch. Chem. Ind. KoZZoide, 1912, 11, 102) that there was shrinkage in volume and corresponding increase in density when caseinogen formed colloidal solution.Ludekings (Wied. Ann., 1888, 35, 552) stated that gelatine in the dry state, with a density of 1.412, gave a value of 1.9 when cdculated from jellies (14 to 35 per cent. gelatine). I n addition to caseinogen, crystalline serum albumin, serum globulin (pseudoglobulin), and crystalline egg albumin, were investigated.The results obtained were as follows : Protein. Caseinogen . . . Egg albumin ... Serum albumin Serum globulin I In Solution. I Dry State. Concentration of Protein per Cent. by Weight. 7.85 14-60 22-15 15.33 I 1.0241 1.0401 1.0647 1.0428 1.390::: 1-359 1-378 1-365 Density of Protein. 1.318 1-269 1.275 1.279 * Corrected for presence of sodium.280 ABSTRACTS OF CHEMICAL PAPERS From the first and second columns of figures may be calculated the exceas degreee of gravity (water 1,000) due to 1 grm.of protein per 100 c.c., and if this be done, the values are 3.07, 2*74,2-92, and 2.79. Figures are also given showing the variation in density of caseinogin (sodium caseinogenate) solutions of different concentrations, the value rising from 1.412 in 9-39 per cent.solution to 1.465 in 1 per cent. solution. The density in solution of these proteins is thus from 5 to 8 per cent. in excess of their densities in the dry state, showing the extent of shrinkage in volume taking place on entering colloidal solution. H. F. E. H. Estimation of Water in Bleached Shellac. H. Wolff. (Farbenzeit., 1913, 18, 1389-1390 ; through Chem.ZentraZbZ., 1913, I., 1549.)-A large proportion of water depresses the melting-point of bleached shellac to such an extent that a sample may fuse even at 60' C. Under these conditions the mass will retain the water, so that long-continued heating (with risk of loss of volatile constituents) at 100' C. is required to expel it completely. For an accurate estimation of the water, the shellac is divided into very small fragments, and 3 to 5 grms.of an average sample are rapidly pulverised and allowed to stand for several hours in a tared dish. This is next heated at about 50" to 60' C. (care being taken that the mass does not fuse) for about an hour, after which the temperature is gradually raised (about 5" C. in thirty minutes) to the melting-point, and is finally maintained somewhat below 100' C.for fifteen to thirty minutes, then cooled and weighed. C. A. M. Estimation of Rosin in Soap. A. A. Besson. (Chem. Zeit., 1913, 37, 453-454.)-To ohtain accurate results in the estimation of rosin by Twitchell's method, the esterification must be^ repeated. From 2 to 2.5 grms. of the fatty acids from the soap are dissolved in 20 C.C.of absolute alcohol, the solution saturated with dry hydrochloric acid gas, and the flask closed and allowed to stand for thirty minutes to an hour. A little pumice and 150 C.C. of water are then introduced, and the liquid boiled until there is a sharp separation of the esters and rosin acids, after which the contents of the flask are transferred to a separating funnel and shaken with ether in the presence of a little methyl orange, and left for two hours.The acid layer is now drawn off, and the ethereal solution shaken with 100 C.C. of alcoholic alkali (potassium hydroxide, 5 grms. ; alcohol, 50 C.C. ; water, 500 c.c.), with phenolphthalein as indicator, and the alkaline extract run off after two hours, this extraction being repeated.The united alkaline extracts are acidified (methyl orange as indicator), shaken with 100 C.C. of ether, and allowed to stand for two hours, after which the aqueous layer is run off, and the ethereal solution evaporated in a tared Erlenmeyer flask. The residue is weighed and mixed with 2 grms. of fatty acids, the behaviour of which in Twitchell's test is known, and the mixture dissolved in 20 C.C.of alcohol, esterified and extracted with ether as described. After removal of the acid layer, 100 C.C. of water are added, and the aqueous layer neutralised with alkali, with methyl orange as indicator, and left for two hours, after which it is run off. Finally, the ethereal solution is titrated with or alcoholic potassium hydroxide solution, with phenolphthalein as indicator. If great accuracyORGANIC ANALYSIS 281 is required, the extraction with alkali must be repeated again, the results being then about 0.3 per cent.higher. C. A. M. Determination of Sucrose in Cane Molasses by the Double Polarisation Method, using Invertase and Acid as Hydrolysts. J. P. Ogilvie. (Intenz. Sugar J., 1912, 14, 89-93.)-1n a previous paper (ANALYST, 1911, 36, 120) on the determination of sucrose in beet products, it was shown by the author that if the selective hydrolyst invertase be used to effect inversion, the results are dis- tinctly higher than those obtained by the ordinary procedure, in which the direct polarisation is taken in a soluti~n that is alkaline owing to the presence of the basic lead acetate used for defecation.It was pointed out that if the direct polarisation be taken in an alkaline solution, then the difference between the direct and inversion readings is not due solely to sucrose, but to some extent to the presence of optically active non-sugars, such as asparagine, aspartic acid, glutamine, and glutamic acid, which have a lower rotation in alkaline than in acid solution, results that are too low to the extent of 0.5 to 1.5 per cent.thus being obtained. Hence, so as to insure that the difference between the direct and inversion readings shall be entirely due to the sucrose hydrolysed, the direct as well as the inversion reading must be taken in an acid solution, and it was shown that when this is done the results are in accordance with those found by invertase, which only hydrolyses the sucrose (and raffinose also, if present) without affecting the optically active amino-acids. The investigation has now been extended to cane molasses, using the same method of procedure, and especially comparing the results found by the invertase method with those given by acid hydrolysis when the sucrose is calculated from the alkaline and acid polarisations.Operating upon Cuban, Egyptian, and Javan molasses, and upon ‘6 American syrup,” results were obtained which lead the author to the following conclusions : (1) That if in the acid hydrolysis method the generally employed procedure of using the alkaline direct polarisat ion be adopted, the results may be too high to the extent of 0.4 to 1.3 per cent., according to the impurity of the product.(2) That if in the acid hydrolysis method the neutral direct polarisation in place of the alkaline be used, the results are nearer the truth, but may again be too high, especially when much reducing sugars are present. (3) That if in the acid hydrolysis method the hydrochloric acid direct polarisation, as recommended by Andrlik, be used, then correct results are obtained.It is further pointed out that whereas in the case of beet molasses it is the optically active amino-acids that vitiate the polarimetric readings, with cane molasses it is the reducing sugars, which are always present to a more or less extent. In the presence of basic lead acetate the dextro-rotation of a cane molasses solution is considerably increased, owing to the effect of the reagent on the laevulose; while after inversion, the solution being acid, this action is eliminated, and the opposite effect obtains-namely, the Iavo-rotation is increased.Thus the difference between the direct and the inversion readings is not solely due to hydrolysed sucrose, and an important source of error is introduced. Concluding, it is pointed out that both readings must be made under the same conditions of acidity by using the hydrochloric282 ABSTRACTS OF CHEMICAL PAPERS or sulphurous acid methods.process is very close, and well within the limit of experimental error. When this is done, the agreement with the invertase J. P. 0. Use of Sulphurous Acid in making the Acid Reading in the Double Polarisation Method of Determining Sucrose. J.P. Ogilvie. (Intern. &w J., 1912, 14, 624-630.)-1n reply to criticisms on the sulphurous acid process of making the acid direct reading in the double polarisation (Clerget-Herzfeld) method of deterlnining sucrose in sugar-factory products, it is pointed out by the author that Pellet's original procedure does not give correct results, and that a modification of it must be used to obtain figures equal to those found by hydrochloric acid (Andrlik).This modification is carried out by saturating the 50 C.C. of defecated solution con- tained in a 100 C.C. flask with sulphur dioxide, or in making up the volume to 100 C.C. with an aqueous sulphurous acid solution containing about 7.5 grms. of sulphur dioxide in 1OOc.c. Figures aregiven indicating that the results thus found are equal to those obtained when hydrochloric acid is used, and that when a lower concentra- tion of sulphurous acid solution than that specified is used, the results do not agree with those obtained with hydrochloric acid. I t is claimed that this method of working with sulphurous acid in excess solves the problem of making the acid direct reading in the case of beet molasses (see previous abstract), and obviates the error due to optically active substances other than sucrose.Moreover, in the case of hydrochloric acid and urea (Andrlik) the rate of inversion is comparatively rapid, with sulphurous acid it is reasonably slow, and the readings remain quite constant in the case of beet molasses for about thirty minutes. For the determination of sucrose in beet molasses by the double polarisation method, using the modified sulphurous acid direct reading, the procedure now recommended is as follows: 200 C.C. of a normal sugar weight solution of the sample, defecated with basic lead acetate and filtered, is prepared. For the direct reading 50 C.C. of this solution are transfetred to a 100 C.C. flask, saturated with sulphur dioxide, the flask being immersed in cold water, and made up to volume with water at standard temperature. Or, instead of saturating the assay liquid with sulphur dioxide, the 50 C.C. may be made up to 100 C.B. with a freshly prepared solution of sulphurous acid, as mentioned above. After well mixing, the liquid is treated with sulphur dioxide, and after adding a little decolourising carbon if necessary, it is filtered, and polarised at 20" C. in a 200 mm. water-jacketed tube. For the inversion reading, 50 C.C. of the same defecated and filtered normal solution of the molasses as that used for the direct reading are transferred to a 100 C.C. flask, 25 C.C. of water, and 5 C.C. of concentrated hydrochloric acid (sp. gr. 1-19> added, the liquid inverted according to the Herzfeld procedure, and read in the same 200 mm. water- jacketed tube at exactly 20" C. Finally, the percentage of sucrose is calculated from the two readings in the ordinary way, applying, however, the constant from the Herzfeld table appropriate to the particular concentration used. Under these con- ditions it is unnecessary to take account of the volume of the lead precipitates produced in making the direct and inversion readings ; but it is pointed out that it is advisable in operating the inversion polarisstion, especially in examining veryINORGANIC ANALYSIS 283 impure products, to carry out the determination in duplicate, continuing heating at 69" C. in the second test for five minutes longer than in the first. I n this way com- plete hydrolysis of the sucrose in the presence of an abnormally high soluble ash and non-sugar content is insured. J. P. 0.

ISSN:0003-2654    

DOI:10.1039/AN9133800271

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

INORGANIC ANALYSIS 283 INORGANIC ANALYSIS. Estimation of Free Acid and Base in Aluminium Sulphate. W. Iwanow. (J. Russ. Phys. Chem. Ges. 1913 45 57-66; through Chem. Zentralbl. 1913 I., 1460-1461.)-The methods generally recommended for the estimation of the free acid or base in aluminium sulphate give inaccurate results. The author therefore pro-poses a method based on the fact that the neutral salts of aluminium are precipitated by potassium ferrocyanide on warming whilst the acid remains in solution and can be estimated by titration. One grm. of the sample to be tested is dissolved in 50 C.C. of water; the solution is heated to 85" C. and 12 C.C. of a solution of potassium ferrocyanide (1 12) are introduced in a rapid stream ; then 20 C.C. of a solution of barium chloride (1 10) are added.After stirring the liquid is transferred to it 100 C.C. flask and diluted to the mark; the precipitate is allowed to settle and 25 to 50 C.C. of the clear solution are withdrawn for titration with Fn sodium hydroxide in presence of methyl orange. If the aluminium sulphate is basic 2 C.C. of hydrochloric acid are added to the original solution and this quantity is deducted from the result found in the titration. Test-analyses made have shown that the proposed method affords useful results. For rapid work the decantation of the clear liquid may be omitted and the whole may be titrated directly using as an external indicator methyl orange paper on which a drop of the solution is placed from time to time during the titration; the results thus obtained however are about 0.2 per cent.too low. J. F. B. Behaviour of Salts of Casium and Rubidium towards Salts of Heavy Metals. M. Wagenaar. (Pham. Weekblad. 1913 50 273-280 ; through Chcm. Zentralbl. 1913 I. 1362-1363.)-IVhen solutions containing cssium or rubidium salts are mixed with solutions of various heavy metals characteristic precipitates are obtained in certain cases and it is shown that a solution of caesium chloride is a useful reagent for the microchemical detection of antimony tin bismuth copper, cadmium aluminium chromium cobalt and zinc. Rubidium chloride may be used for the detection of antimony tin bismuth copper aluminium and chromium. w. P. s. Estimation of Calcium as Tungstate. A. Saint-Sernin. (Compt. rend., 1913 156 1019-1020.)-When excess of a 20 per cent.solution of normal sodium tungstate is added to a boiling ammoniacal solution of calcium chloride the calcium is quantitatively precipitated as CaWO,. As the corresponding salt of magnesium is easily soluble the method affords a ready means of separating calcium from magnesium. As the precipitate is crystalline and very heavy (sp. gr. 6-05) it is easily washed by decantation. For the estimation of calcium it must be collecte 284 ABSTRACTS OF CHEMICAL PAPERS on a tared filter and dried at 100" C. as burning the paper even after detaching the precipitate as far as possible always leads to low results. No test numbers are given. G. C. J. Analysis of Hydrosulphites. E. Bosshard and W. Grob. (Chem. Z e k 1913 37 423-425 ; 437-439.)-After reviewing various methods which have been proposed for the analysis of hydrosulphites the authors recommend a modification of the process in which the salt is titrated with ammoniacal copper sulphate solution.The procedure is as follows A weighed quantity of about 0.1 grm. of the salt is placed in an Erlenmeyer flask and a current of pure hydrogen is passed into the latter. After twenty minutes and when all the air has been expelled from the flask, ammoniacal copper sulphate solution is run in from a burette the jet of which passes through the rubber stopper closing the flask. The quantity of copper solution added at first is a few C.C. short of that required to decompose the hydrosulphite completely (that is to produce a blue coloration in the flask) ; 9 drops of a 0.1 per cent.indigo-carmine solution are then added and the titration with the copper solution is continued until the yellow colour of the contents of the flask changes to light blue. The copper solution is prepared by dissolving 4.469 grms. of crystallised copper sulphate in boiled and cooled water adding a large excess of ammonia and diluting the whole to 1 litre; the actual quantity of copper per C.C. of this solution must be estimated electrolytically. The reaction proceeds according to the equation : 2CuS0 + Na,S20 + 4NH + 2H,O = Cu,SO + Na2S0 + (NH,),SO + (NH,)2S0,. The results found are affected to a slight extent when bisulphites are present in the sample in considerable proportion the figures obtained being about 0.3 per cent. too high. Hydrosulphites may also be estimated by an acidimetric method which is based on the reaction : Na2S,0 + H20 + O,= NaHSO + NaHSO,.The two hydrogen salts formed are titrated with barium hydroxide solution, using phenolphthalein as indicator. The operation is carried out in an atmosphere of hydrogen and the results obtained agree with those yielded by the copper titra-tion method. Processes which depend on the reduction of mercuric chloride by hydrosulphites yield fairly accurate results and may be used for technical purposes. W. P. S. Monohydrated Ferrous Sulphate and its Employment in Volumetric Analysis. D. Florentin. (Bull. SOC. Chim. 1913 13 362-366.)-Monohydrated ferrous sulphate is recommended as a suitable standard for the control of solutions of permanganate. I t is easily prepared is not hygroscopic does not oxidise appreciably on storage in a stoppered bottle and neither loses water nor suffers oxidation when heated €or several hours at 120" C.To prepare it commercial ferrous sulphate is first purified by two or three disturbed crystallisations from very faintly acid solution. Some 400 grms. of the purified salt are then placed in a 500 C.C. flask together with 200 C.C. of 50 per cent (by weight) sulphuric acid and heated on a water-bath. Th INORGANIC ANALYSIS 285 salt begins by dissolving almost completely ; then suddenly on the attainment of a certain temperature a white crystalline precipitate appears. Heating is continued for a few minutes and the precipitate is then brought on to a Buchner funnel, washed with 96 per cent.alcohol or anhydrous acetone then with anhydrous ether, and finally dried in a vacuum desiccator. The yield is about 40 grms. The salt contains 32.93 per cent. of iron (theory for FeSO,.H,O = 32-94 per cent.) and is free from any but negligible traces of ferric salt. On exposure to the air of the laboratory for fifteen days 3 grms. appeared to increase in weight 1.5 mgrms. and 4 grms. lost 1.5 mgrms. when kept for four hours at 120a C. G. C. J. Estimation of Total Carbon in Steel and Ferro-Alloys by Combustion in Oxygen under Pressure. W. E. von John. (Chem. Zeit. 1913,37 426-427.)-An apparatus of simple construction is described for this purpose. The weighed portion of the sample is placed in a magnesia crucible supported in an Erlenmeyer Bask by means of a wire attached to a copper tube which passes through the rubber stopper of the flask.Two electrodes also pass through the stopper and their lower ends extend into the crucible. At the commencement of an estimation the flask is filled with oxygen a known quantity of & sodium hydroxide is introduced and the rubber stopper bearing the electrodes and copper tube with the crucible and its con-tents is inserted. The upper end of the copper tube is connected with a reservoir of oxygen under slight pressure; a bottle containing mercury is placed between the reservoir and the combustion flask and serves as a safety-valve. At the end of the combustion which is carried out under a pressure of aboat 1 atmosphere the excess of sodium hydroxide in the flask is titrated using phenolphthaleln as indicator and the quantity of carbon present is calculated from the carbonate formed.The presence of sulphur (up to 0.8 per cent.) and of phosphorus (up to 2 per cent.) in the steel or iron alloy does not affect the accuracy of the method. w. P. s. Colorimetric Method for Determination of Chromium in Steel. F. Garratt. (J. Ind. a d Eng. Chem. 1913 5 298-300.)-The author finds that Koenig’s method (ANALYST 1911 36 238) is better adapted than any other for the estimation of small quantities of chromium in steel from 0.6 per cent. down to the smallest quantity that can have any metallurgical significance. The method is applied to steel as follows From 0.2 to 0.4 grm. (the larger quantity only for steels with less than 0.15 per cent. chromium) is dissolved in 10 C.C.of dilute (1 3) sulphuric acid 0.5 C.C. nitric acid is added to oxidise the iron and the solution is evaporated until fumes of sulphuric acid appear. To the cooled and slightly diluted solution 50 C.C. of 10 per cent. sodium hydroxide are added together with about 1 grm. of sodium peroxide. The mixture is then boiled for five minutes to destroy hydrogen peroxide cooled diluted to exactly 200 c.c. and filtered. To 100 C.C. of the filtrate 2 C.C. of 85 per cent. phosphoric acid 8 C.C. of concentrated sulphuric acid and 2 C.C. of Koenig’s reagent (a 1 per cent. solution of disodium 1.8-dihydroxy-naphthalene 3.6-disulphonate) are added and the pink colour is compared with that produced by treating an equal weight of chromium-free steel and a measured volume of standard bichromate solution in a similar manner.As about 5 per cent. of th 286 ABSTRACTS OF CHEMICAL PAPERS chromium present is carried down with the precipitate of ferric hydroxide steel must be used in the blank test. Neither tungsten molybdenum nor titanium inter-fere. Vanadium interferes by imparting a brownish tint and where the amount of vanadium equals or exceeds the amount of chromium the results are worthless I n the commoner case where the chromium preponderates the results are too high by about one-third of the vanadium present. Thus a steel with 0.2 per cent. chromium and 0.1 per cent. vanadium appeared to contain 0.23 per cent. chromium when tested as described and the paper includes a number of analyses showing that the error due to vanadium closely approximates one-third of the vanadium present.When the ratio of chromium to vanadium is high this error becomes less than the unavoidable error of a colorimetric method and in other cases it can be reduced to negligible proportions by the use in the blank test of a vanadium steel of approxi-mately the same vanadium content as the samples. G. C. J. Estimation of Manganese in Castings and Steels by the Method of G. von Knorre. M. Huybrechts and N. Joassart. (BzcZZ. SOC. Chim. de Belg., 1913 27 130-137.)-The authors have developed a modihation of von Knorre’s method for the separation of manganese from large quantities of iron depending on the decomposition of manganese persulphate by heating and its precipitation as the dioxide. The authors confirm the observation that the precipitation is very unsatis-factory when any considerable quantity of acid is present and have taken steps to deal with this difficulty.A complete separation of the iron is not required but merely a concentration of the manganese in a state in which its estimation may be effected by the Guyard-Volhard or other suitable method. The procedure adopted in the case of manganiferous metals is as follows Three to four grms. of the metal are dissolved in the least possible excess of nitric acid (11 C.C. of acid of 7 concentra-tion per grm. of srtmple) in an Erlenmeyer flask with 40 to 50 C.C. of water the liquid being heated towards the end of the operation. The solution is diluted with 250 C.C. of water treated with 50 C.C. of an 8 per cent.solution of ammonium per-sulphate boiled for fifteen to twenty minutes removed from the flame again treated with 50 C.C. of the persulphate and 50 C.C. of an 8 per cent. solution of sodium carbonate added together and again boiled for twenty minutes. The precipitate is filtered off the filtrate being returned until clear ; the precipitate is not washed but merely drained. A deposit remains on the walls of the empty flask; this is dissolved by running 3 to 4 C.C. of strong hydrochloric acid down the sides of the flask while the latter is still warm. The filter containing the precipitate is placed over the same flask a hole is made and the precipitate is washed into the acid by a jet of boiling water. Meanwhile a mixture of 15 C.C. of hydrochloric acid and 15 C.C.of hydrogen peroxide is prepared and poured drop by drop on the filter ; in this way the last residues of precipitate are dissolved on the filter and washed into the flask with boiling water. The solution is boiled to destroy the hydrogen peroxide and is then diluted to a definite volume and filtered to remove paper fibres. The solution contains all the manganese and a little of the iron; aliquot portions of this are taken for the determination of the manganese by the Guyard-Volhard method. Zinc oxide is added in quantity just sufficient to precipitate the whole of the iron the additio INORGANIC ANALYSIS 287 being stopped when the precipitate of ferric hydroxide is reddish-brown rather than yellowish-brown in colour; an excess of zinc oxide should be avoided.The manganese is titrated at a temperature as near 100” C. as possible. The filtrate from the manganese precipitate should be tested colorimetrically to ascertain whether any of the manganese has escaped precipitation. It is diluted to 500 c.c., and 50 c.c are placed in a 100 C.C. flask. To this are added in succession 3 grms. of potassium dihydrogen phosphate 10 C.C. of nitric acid (sp. gr. 1-2) and 5 C.C. of TG silver nitrate solution. The mixture is heated on the water-bath for ten to fifteen minutes 3 grms. of ammonium persulphate dissolved in a little water are added and the solution is shaken cooled and diluted to 100 C.C. The colour of the liquid is then compared with standards prepared from permanganate solution. If the acidity of the original solution from which the manganese was precipitated has been sufficiently controlled this colour test should show no manganese in the filtrate.J. F. B. Direct Estimation of Elementary Nitrogen by Means of Commercial Calcium Carbide. B. Natus. (Zeitsch. anal. Chem. 1913 52 265-292.) - The method depends on the fixation of the nitrogen as calcium cyanamide and its subse-quent determination by subjecting the calcium cyanamide to Kjeldahl’s process. I n the illustration AA is a porcelain tube 30 cm. long by 15 mm. wide ; g’ g” are 6 cm. lengths of hard glass rod to reduce the gss-space; S is a porcelain boat 5 to 6 cm. long; B is an asbestos mardle wrapped round the asbestos rings CC. The mantle has a large opening D for the admission of the hot products of combustion from the burner H to the space between the tube and the mantle and a series of small holes aa near the end of this space for the escape of the hot gases secures a more uniform heating of the boat.The combustion-tube is covered by tiles not shown in the illustration and the rubber stoppers F’ F” are protected by perforated asbestos discs slipped over the tube also not shown. E’ E” are capillary tubes and L M N are gas burettes I m n being their cocks and b b b rubber connections. P connects to a supply of dry hydrogen. The boat is charged with about 2 grms. of a finely powdered mixture of calcium carbide (10 parts) and fused calcium chloride (1 part), ignited at a red heat in a current of dry hydrogen allowed to cool in the gas 288 ABSTRACTS OF CHEMICAL PAPERS weighed and transferred to the apparatus.I t is necessary to weigh the carbide mixture as even after ignition it always retains nitrogen (1 to 2 C.C. per grm.). Each batch of 100 grms. is tested by igniting about 2 grms. as described and submitting the residue to Kjeldahl’s process; a correction is then applied to all results obtained with this mixture. The boat plugs g‘ g” stoppers 3” F” and capillaries E’ E” being in place but the burettes disconnected f is connected to a supply of dry hydrogen, which is passed through the apparatus for fifteen minutes d being plugged after the first few minutes. Meanwhile L is charged with the gas mixture and after fifteen minutes the three burettes M and N being full of mercury are connected with their cocks closed to the capillaries Some 30 C.C.of hydrogen are run over into N the cock m again closed the stream of hydrogen stopped and f plugged. After thirty minutes the volume of gas in L its temperature and the barometric height are noted and the combustion-tube raised to 900’ to l,OOOo C. after :opening n so that the hydrogen may expand into N. A higher temperature than l,OOOo C. is to be avoided lest calcium cyanamide be partially decomposed or sublimed. When the mercury level in N no longer sinks from 12 to 15 C.C. of the gas mixture in L are driven into the apparatus and n is closed. The absorp-tion is rapid at first (over 10 C.C. per minute) but gradually becomes slower. If the mixture in L is mostly nitrogen the whole of it can be gradually transferred to the apparatus by manipulating the levelling-tube of L so that the pressure is but slightly above that of the atmosphere Cock I is then closed and m opened when, if the gas mixture being examined is mainly nitrogen as much as 10 C.C.of hydrogen will gradually enter the apparatus to take the place of the absorbed nitrogen. When the level of the mercury in M no longer rises the attached levelling-tube is raised, n opened and the gaseous contents of the apparatus slowly driven over into N and then back into M. This operation may be repeated when mercury is used as con-taining liquid. The apparatus is now allowed to cool and the boat with its contents, which adhere somewhat firmly to it is introduced carefully into a Kjeldahl flask and 25 C.C. of 98 per cent. sulphuric acid added so as to wet the sides of the flask without touching the contents of the boat which should float.The flask is then agitated so that the contents of the boat are thoroughly and instantaneously wetted. A globule of mercury is added and the mixture heated to boiling till colourless (one hour); it is then cooled made alkaline treated with potassium sulphide and distilled into standard acid using zinc dust to prevent bumping. Importance is sttached to the use of zinc for this purpose and to the manner of bringing the material in contact with acid. Different procedure gave erratic results. Working as above the extreme error (always negative) of the method can be kept below 0.3 C.C. nitrogen. The author’s experiments were made with about 50 C.C. of nitrogen. If water be used as containing liquid a slight additional correction must be made for nitrogen which may be found in hydrogen even when this gas is carefully prepared and stored over boiled water; but the chief error introduced by the use of water is due to the partial decomposition of calcium cyanamide by water-vapour.When using water therefore the sweeping of the gas back and forth from M to N should not be repeated more than twice. With this precaution the error (always negative) can be kept below 0.9 C.C. G. C. J INORGANIC ANALYSIS 289 Detection of Nitrous Acid in the Presence of Ferric Salts. P. Altmann. (Ghem. Zeit. 1913 37 501.)-Ferric salts do not interfere with iodimetric estimations provided that the solution is nearly saturated with a phosphate and that an excess of potassium iodide is avoided.Advantage may be taken of this fact for the iodimetric estimation of ammonia in the presence of ferric salts. I t is also possible in this way to detect nitrous acid in the presence of ferric salts. Eight grms. of pure disodium hydrogen phosphate are dissolved in 100 C.C. of water containing ferric salts and the solution shaken with about 0.2 grm. of potassium iodide until only a white turbidity of ferric phosphate is perceptible. The liquid is then acidified with 5 C.C. of 2 sulphuric acid and 2 C.C. of starch solution added a blue coloration being a t once produced in the presence of a trace of a nitrite. If large quantities of ferric salts are present the speed of separation of the iodine by the nitrous acid is accelerated. By this test it is possible to detect from 0.1 to 0.2 mgrm.of nitrous acid in the presence of 500 mgrms. of ferric iron per litre. C. A. M. Precipitation of Palladium by Means of a-nitroso-P-naphthol. W. Schmidt. (Zeitsch. anorg. Chem. 1913 80 335-336.)-When 3 drops of a saturated solution of a-nitroso-P-naphthol in 50 per cent. acetic acid are added to 1 C.C. of a solution containing a trace of palladium a voluminous red-brown precipitate results. The reaction is shown even by solutions containing only 1 part of palladium in 1,000,000 if some little time be allowed and none of the other platinum metals give a similar reaction. Applied to the quantitative determination of palladium in pure solutions and in amounts such as 0.02 grm. the results were exact SO far as could be shown by a balance sensitive only to 0.1 rngrm.I n presence of 1 grm. of platinum and 0.2 grm. rhodium 0.0190 grm. palladium was returned when 0.0188 grm. was present so that the method appears suited to the estimation of palladium in presence of much platinum and a considerable amount of rhodium. The solution to be examined is rendered faintly acid with nitric acid heated to boiling the reagent added the precipitate filtered and washed with hot water the precipitate ignited with the paper in a porcelain crucible and the residual palladium weighed. I t is shown that the precipitate has the composition (C,,H,NO. O),Pd. G. C. J. Detection of White Phosphorus in Match Compositions. F. Schroder. (Arbeit. Kaiserl. Gesundheitsamte 1913 44 1-29.)-The experimental part of this paper is a continuation of the work of A.Siemens (Arbeit. Kaiserl. Gesundheitsamte, 1906 24 264). A table of the substances used in the match industry is given; of these red and scarlet phosphorus phosphorus sulphides ‘‘ sulphophosphite ” (zinc polyhypothiophosphite) phosphorus suboxide and solid hydrogen phosphide are alone likely to contain white phosphorus. The usual tests for phosphorus are not specific for the white variety ; the test should moreover be insensitive to the small quantities unavoidably left in the commercial product. Siemens (Zoc. cit.) has given a method applicable to red phosphorus for which the Mitscherlisch test fails. It is also useless for the phosphorus sulphides except diphosphorus pentasulphide. Experiments on the phosphorescence of the sulphides pure or contaminated with white phosphorus 290 ABSTXACTS OF CHEMICAL PAPERS led up to the following method Strips of Schleicher and Schiill No.597 filter-paper 3x10 cm. are saturated with 0.5 C.C. of a solution made by dissolving 3 grms. of the phosphorus sulphide sulphophosphite red or scarlet phosphorus, Or match composition in 150 C.C. benzene or in the case of very sensitive com-positions 200 match heads in 15 C.C. of benzene. The strips are suspended with a thermometer in a vertical glass tube 50 x4-2 cm. and are heated to 45"-50" C. by a current of air rising through the tube from the interior of closed Victor-Meyer water-bath beneath. The temperature must not in any case exceed 55" C. Screens are arranged to cut off all light from the burner and the whole apparatus is set up in a dark room.With the quantities given no phos-phorescence should be observed as the benzene evaporates; but if not less than 0.02 mgrm. or in the case of tetraphosphorus trisulphide 0.04 mgrm. of white phosphorus is present in 1 C.C. of the benzene solution the paper becomes dis-tinctly phosphorescent ; with more the intensity is increased and luminous clouds appear. According to the complete scheme of examination appended 1 C.C. of the benzene solution is first tested with 1 C.C. of a solution containing 1.7 grms. silver nitrate in 100 C.C. of ammonia of sp. gr. 0.992. If on shaking and allowing to separate the aqueous layer is unaltered or tinted yellow no white phosphorus is present. If it is coloured red or brown or contains a brown or brown-black preci-pitate the rest of the benzene solution must be subjected to the phosphorescence test unless the substance under examination is red phosphorus when the presence of white phosphorus is proved and further testing is unnecessary.If required the presence of phosphorus sulphides is determined by testing the benzene solution with lead nitrate If the test with ammoniacal silver solution is positive and the phosphoresceuce and lead nitrate tests negative scarlet phosphorus is present. 0. E. M. Determination of Phosphorus in Ferrotungsten Metallic Tungsten Powder Tungsten Oxide and Tungstic Acid by Direct Solution. C. M. John-son. (J. Ind. and Eng. Chem. 1913 5 297-298.)-In the author's experience the practice of fusing tungsten-bearing materials with sodium carbonate and nitre, leaching out the fu~ion removing tungstic acid by evaporation with acid and using the filtrate from tungstic acid for the determination of phosphorus leads to the under-estimation of that element.I n no case has he succeeded in recovering as much as one-third of the phosphorus present and in some cases less than one-tenth was recovered the rest being carried down as phosphotungstic acid with the tungstic acid. The only method he has found sufficiently accurate for technical purposes is the following somewhat lengthy one : Tungsten ores or tungsten oxide (1 grm.) in the finest possible state of division is extracted for at least six hours at a nearly boiling temperature with 100 C.C. of con-centrated hydrochloric acid adding about 0.1 grm.of potassium chlorate every half-hour or so. When decomposition is complete the yellow tungstic oxide will show no trace of the brown or grey colour of the original ore. The mixture is then evaporated to dryness the residue heated for ten minutes with 50 C.C. of hydrochloric acid, 50 C.C. of water are added the mixture is boiled for fifteen minutes cooled and afte INORGANIC ANALYSIS 291 addition of some paper pulp filtered through a double filter. The precipitate is washed with dilute (1 20) hydrochloric acid and the filtrate and washings treated in the manner described in the next paragraph but one. Tungsten powder is first roasted at a low red heat until converted into the yellow oxide; it is then treated as above described. Ferrotungsten (I grm.) is decomposed by treatment with 30 C.C.of concentrated nitric acid with the slow addition of 3 c.C. of hydrofluoric acid followed by warming and finally boiling until a clear solution is obtained. A porcelain dish can be used but a little more hydrofluoric acid must then be taken and a blank experiment run using a standard steel as there is a tendency to get high results when using porcelain. The solution is evaporated t o dryness but not baked; 50 C.C. of hydrochloric acid are added and the solution is again evaporated to dryness ; the residue is heated with 20 C.C. of hydrochloric acid 50 c.C. of water are added and the solution filtered from the bulk of the tungstic acid. To separate the last of this the filtrate and washings for which dilute (1 20) hydro-chloric acid is used are evaporated to 10 c.c.diluted to about 30 c.c. and refiltered. The filtrate end waghings are evaporated to 10 c.c. 75 C.C. of nitric acid are added, and the mixture heated with a cover on until all action is over. I t is then evaporated to 20 c.c. 50 C.C. of nitric acid are added and evaporation is continued until only about 15 C.C. remains. About 20 C.C. of water are then added the liquid filtered and the filter washed fifteen to twenty times with 2 per cent. nitric acid. The filtrate and washings are concentrated to 40 c.c. excess of permanganate added the mixture boiled three to four minutes manganese dioxide dissolved by means of ferrous sulphate and the phosphorus precipitated by means of molybdate solution. G. C. J, Estimation of Phosphorus in Iron and Steel.P. Artmann and J. Preisinger. (Zeitsch. angezo. Chem. 1913 26 203-208.)-Some sixsy methods which have been proposed for this estimation are discussed the opinion expressed that the most accurate of them is probably that which depends on collecting the phosphomolybdate precipitate on a tared Gooch filter heating to 160' to 180' C. and weighing and the following method which is shown to give results in close accord with those obtained by the foregoing method is recommended on the ground of sim-plicity. In essentials it is identical with the method of Artmann and Brandis (Zedsch. anal. Chena. 1910 49 l) and depends on the estimation of the ammonia in the phosphomolybdate precipitate by treating this with a measured excess of standard hypobromite solution and determining the excess of hypobromite iodi-metrically.Applied to iron or steel the method is conducted as follows The sample (0.3 to 1 grm. cast iron or 2 to 4 grms. steel) is dissolved in 25 to 100 C.C. of nitric acid (sp. gr. 1*135) and after three minutes the solution is evaporated to half bulk. I n the case of cast iron graphite is filtered off and washed with nitric acid (sp. gr. 1.135). With steel 10 C.C. of the acid are added and then in either case 5 C.C. of 5 per cent. permanganate. The solution is boiled till the per-manganate colour disappears 5 to 10 C.C. of 3 per cent. hydrogen peroxide are added to reduce manganese dioxide and the solution boiled till clear. The volume of the solution is now increased to 50 to 100 C.C.by addition of nitric acid (sp. gr. 1-135) 292 ABSTRACTS OF CHEMICAL PAPERS 5 to 20 grms. ammonium nitrate are added the mixture heated to boiling and then withdrawn from the source of heat. To the solution still at about 9 5 O C. 40 to 60 C.C. of molybdate solution are added slowly so that the addition occupies about three minutes the mixture being stirred vigorously the while and for a further five minutes. The molybdate s o h tion contains 75 grms. ammonium molybdate and 110 C.C. of concentrated nitric acid per litre. Tho liquid now at about 40" C. is decanted through a filter and the precipitate is washed twice by decantation with a 5 per cent. solution of ammonium nitrate containing 0.5 per cent. nitric acid then trans-ferred to the filter and washed three or four times with the same mixture and finally five or six times with 8 to 10 C.C.of water. The precipitate and paper are now intro-duced into a mixture of 10 C.C. standard hypobroinite and about 20 C.C. water. The hypobromite solution is about F and is made by adding gradually 1,400 C.C. of bromine water to 500 C.C. of well-cooled sodium hydroxide. Sodium phosphate (7 grms.) is next added to inhibit the reduction of molybdic acid by hydrogen iodide, water is added to bring the volume up to about 80 c.c. the mixture agitated to dissolve the phosphate 0.5 grm. potassium iodide and 10 to 15 C.C. of sulphuric acid are added and the liberated iodine titrated with & thiosulpbate. The difference between the volume of thiosulphate required in a blank test and that required in an analysis conducted as described is the measure of the phosphorus present.P = 3NH3 = 91. Even proportions of phosphorus as small as 0.02 per cent. can be estimated with an error not exceeding 5 per cent. G. C. J. Separation of Phosphoric Acid by Means of Stannic Acid in Quali-tative Analysis. W. Meeklenburg. (Zeitsch. anal. Chem. 1913 52 293-298.)-The author has recently shown (Zeitsch. anorg. Chem. 1912 74 215) that Reynoso's method for the separation of phosphoric acid in the course of qualitative analysis by boiling the nitric acid solution with tin depends on the adsorption of phosphoric acid by metastannic acid. Some skill is required to use the original method success-fully b u t the substitution of ready-made hydrated stannic oxide for tin makes the separation quite easy.The reagent is prepared by adding gradually 150 grms. of granulated tin to 1,500 C.C. of ice-cold dilute nitric acid (sp. gr. 1.2). The colloidal sol is poured into 10 litres of water a t room temperature and the resulting preci-pitate is washed several times by decantation filtered off on linen washed till the washings are nearly neutral and allowed to drain. The gel on the filter which contains about 10 per cent. of stannic oxide constitutes the reagent. As much as 40 grms. is required to adsorb 0.1 grm. of phosphoric acid a t one operation, but the reagent is more efficient if used as described below. The solution to be analysed freed from metals of Groups I. and II. and containing about 15 C.C. of nitric acid in 100 c.c. is treated with a tablespoonful of the gel boiled for fifteen minutes and filtered.The filtrate is treated in a similar manner the first runnings of the second filtrate rejected if cloudy and the subsequent clear filtrate collected. I n it any metals except iron titanium and zirconium which were present in th INORGANIC ANALYSIS 293 original solution can be detected. With as much as 0.7 grm. phosphoric acid present in the original solution less than 1 mgrm. finds its way into the filtrate. If the original solution contains chlorides or large quantities of iron aluminium or chromium some tin may pass into solution. It may be separated together with any copper or lead derived from the original “tin,” by means of hydrogen sulphide, before proceeding with the analysis but in absence of chlorides the necessity for this is rare.Zirconium was not experimented with but it is pointed out that it forms a compound with phosphoric acid which is insoluble even in strong nitric acid. Iron and titanium are wholly retained by the gel and may be tested for by simple treatment of the gel with ferrocyanide and hydrogen peroxide respectively. If traces of these metals have to be looked for it is best to make the first separation with very little of the gel and to filter this off and test it. This concentrates the iron and titanium into a small bulk and enables very small quantities to be detected. G. C. J, Iodimetric Estimation of Persulphuric Acid. E. Muller. (Zeitsch. anal. Chem. 1913 52 299-303.)-When a persulphate is added to excess of a solution of sodium hydroxide and potassium iodide the solution first acquires a yellow colour due to hypoiodite but finally becomes colourless and then contains iodate.The reaction which may be represented by the equation-3K2S20 + K I + 6KOH = 6K2S0 + KIO + 3H,O, proceeds slowly at room temperature but is complete in less than three minutes at 100” C. As it proceeds quantitatively it affords a ready means of estimating persul-phates the solution of iodate with excess of iodide being acidified and the resulting iodine titrated with thiosulphate. I t will be seen that the net result of the reaction between persulphate and iodide and the well-known one between iodic and hydriodic acids is that K,S,O,=I, or that 0.5 grm. of potassium persulphate will give a convenient reading (about 40 c.c.) if -& thiosulphate be employed.The process is conducted as follows t o 25 C.C. of sodium hydroxide contained in a flask with 2 grms. potassium iodide about 0-5 grm. of the salt to be analysed or an equivalent quantity of a solution is added rtnd the mixture boiled gently for three minutes. It is then cooled acidified by addition of 30 C.C. sulphuric acid and titrated with thiosulphate. Sodium hydroxide purified by alcohol is to be avoided as it usually exercises a reducing action on persulphate. If an acid solution is to be analysed it may be necessary to take a larger quantity of sodium hydroxide than is specified above. The necessity will be shown by the immediate separation of iodine but it does not suf5ce merely to decolourise this by addition of alkali; there must be a fair excess of the latter present as the first reaction uses up alkali.The method is unaffected by Caro’s acid which is always present in solutions of persulphuric acid which have been prepared some days. On the other hand it breaks down in presence of hydrogen peroxide which is also liable to be present in such solutions BO that its use is practically limited to the analysis of solid substances and neutral solutions. G. C. J 294 ABSTRACTS OF CHEMICAL PAPERS Estimation of Platinum. L. St. Rainer. (Oesterr. Zeit. Berg. u. Hiittenwesen, 1913 61 141-147 155-160 ; through Chem. Zentralbl. 1913 I. 1542-1543.)-For the estimarjion of platinum in alloys a representative sample of the repeatedly-fused metal and (0.25 grm. to 1 grm.in the case of poor alloys) is weighed with sufficient refined gold and silver to give a ratio of P t Au Ag = 1 10 30. At the same time a control test is made with an equal weight of refined silver which is cupelled simultaneously with the alloy mixture and with an equal quantity of lead flux. The cupellation is started at 900' C. and the heat afterwards raised to 1,200" C. until the operation is finished after which the cupels are left for ten minutes in the opening of the muffle furnace and then cooled to increase the malleability of the buttons. The latter are now weighed and the amount of added gold and silver deducted from the platinum button the quantity of silver being checked by the result obtained in the control cupellation. Both buttons are then rolled in the usual way, and boiled for thirty minutes with sulphuric acid (sp.gr. 1-84> after which the flask is cooled and the boiling repeated for thirty minutes with fresh acid. The residue is washed with dilute sulphuric acid and with water and then digested for several minutes with nitric acid followed by an excess of ammonia or with an ammoniacal solution of ammonium acetate to dissolve any lead sulphate present, after which the rolls now consisting of platinum and gold are dried ignited and weighed and the amount of added gold deducted. Each roll is then alloyed with three times its weight of refined silver and the resulting buttons again made into rolls boiled with nitric acid and subsequently weighed. This treatment is repeated twice to separate the small amounts of platinum retained by the gold Provided the proportions given of silver are present the whole of the platinum will be dissolved after the second treatment.A similar method may be used in the examination of platiaum dross but in that case the control cupellation of silver may be omitted. C. A. M. Cobaltinitrite Method for the Estimation of Potassium. J. L. M. van der Horn van den Bos. (Chem. Weekblad 1913 10 182-186 ; through Chem. Zentralbl. 1913 I. 1362.)-The following procedure is recommended The solution under examination is evaporated and the residue is treated with 3 C.C. of a 50 per cent. cobaltous chloride solution and 3 C.C. of a 50 per cent. sodium nitrite solution, the mixture is stirred thoroughly then acidified with 2 C.C. of 50 per cent.acetic acid and placed aside for forty-eight hours. The precipitate is collected on a filter, washed and then rinsed into a beaker an excess of & potassium permanganate is added the mixture warmed and after the addition of 15 C.C. of sulphuric acid (1 I), the excess of permanganate is titrated with oxalic acid solution. The precipitate has the formula K,NaCo(NO,), and according to theory 1 C.C. of & permanganate should be equivalent to 0*000156 grm. of K,O but experiment has shown that 1 C.C. of & permanganate corresponds with 0*000172 grm. of K,O. w. P. s. New Physic0 - Chemical Volumetrie Method. R. Dubrisay. (Compt. rend. 1913,156 894-898.)-1t was found by Donnan that the superficial tension at the surface of separation of water and a petroleum hydrocarbon was very greatl Ih'ORGANIC ANALYSIS 295 reduced by the presence of a trace of alkali in the water when the hydrocarbon contained a fatty acid such as stearic acid in solution.On this observation the author has based a method of determining the exact point of neutralisation of strong acids. The liquid under examination is introduced into a curved pipette which is dipped into the water upon which there is vaseline oil previously purified and saturated with stearic acid. The point of the pipette is made to coincide with the juncture of the oil and water and the number of drops issuing from the pipette into the oil during the fall of the liquid between two fixed points is then counted. Thus for example the following results were obtained with hydrochloric acid 20 C.C.of which required for neutralisation 18.8 C.C. of & sodium hydroxide solution : Number of C.C. of alkali added Number of drops . . 45 45 46 456 75 98 Similar results were obtained with nitric and sulphuric acids. In the case of boric acid the hydrolysis of the alkali salt was clearly indicated whilst phosphoric acid a t first behaved like a strong acid but soon showed signs of the hydrolysis of the disodium salt and subsequently of the neutral salt. to 20 C.C. of acid . . 0 17.9 18.6 18.7 19.1 19.7 C. A. M. Detection of Potassium by Means of Tartaric Acid. L. W. Winkler. (Zeitsch. angew. Chem. 1913 26 208.)-As is well known this test is apt to fail owing to the tendency of potassium bitartrate to form supersaturated solutions. The superior efficacy of the powdered reagent as compared with a solution is due to the presence of a trace of potassium salt seldom absent from commercial samples of tartaric acid.The author recommends the following modification of the test About 0.5 grm. of sodium acetate is first disnolved in 10 C.C. of the solution to be examined, or in 10 C.C. of a 5 per cent. solution of a solid and 0.5 grm. of powdered commercial tartaric acid is then added and the mixture shaken. I n absence of potassium (ammonium rubidium and caesium) the solution becomes clear at once as the powdered acid dissolves rapidly but 20 mgrms. of potassium will give an immediate precipitate and 10 mgrms. will give one within two minutes. The reaction succeeds equally well with halides nitrate chlorate and sulphate and with salts of organic acids except tartar emetic.Potash alums react somewhat less well. The use of alcohol is apt to make the test less certain as it may throw out other salts than potassium bit artrate. G. C. J. Preparation of Selenic Acid and Sodium Selenate as Reagents for the Estimation of Bromine in Halogen Salts. P. L. Blumenthal. (Zeitsch. anorg. Chem. 1913 80 246-251.)-Gooch and Blumenthal have described a method (ANALYST 1913,227) for the estimation of bromide in presenceof chloride which depends on the use of selenic acid or sodium selenate. When these reagents are prepared by the usual methods (Gmelin-Kraut Handb. d. anorg. Chem. 1907 I. 769) they are apt to retain impurities which interfere with their use for the above purpose.Selenic acid of a satisfactory degree of purity may be prepared as follows Selenious acid is first prepared by dissolving elementary selenium in nitric acid and evaporating to dryness. The crude product is sublimed over manganese dioxide (cf. Norton 296 ABSTRACTS OF CHEMICAL PAPERS Zeitsch. anorg. Chem. 1899 20 222) and 10 grms. of the sublimate are treated with 75 to 100 C.C. water 6 grms. potassium bromate and 4 C.C. nitric acid and the mixture is heated on the water-bath. Bromine is set free and selenic acid results and on evaporating the solution to small bulk the excess of bromic acid decomposes the nitric acid with liberation of more bromine. When this reaction is over and the solution is colourless more bromate is added little by little so long as an addition gives rise to the liberation of bromine showing that some nitric acid is yet undecornposed.When the solution remains colourless on addition of more bromate, it is made faintly alkaline with ammonia heated to boiling and the selenic acid precipitated by means of barium nitrate. The barium selenate is filtered off a t the pump washed with hot water and dried at a red heat. I t is then decomposed by means of 5 C.C. of sulphuric acid 20 C.C. of water is added and the mixture heated for two to three hours on the water-bath diluted filtered and washed until the filtrate and washings amount to about 30 C.C. The resulting solution is a 40 per cent. solution of selenic acid containing about 5 per cent. of selenious acid but free from any impurity which would interfere with its use in the estimation of bromine.Sodium selenate is prepared by mixing 105 grms. of powdered selenium with 500 grms. of sodium peroxide and fusing the mixture in batches of about 50 grms. in nickel crucibles. The cooled melt is dissolved out with water and the solution filtered through asbestos and evaporated as rapidly as possible over a free flame to a syrupy consistency. The bulk of the caustic alkali is then extracted by means of hot 97 to 99 per cent. alcohol and the extraction repeated six or seven times. The residue of sodium selenate and carbonate is dissolved in water and the solution carefully neutralised by means of sulphuric acid and evaporated to incipient crystallisation and allowed to cool. The crystals are separated from the mother liquor which is further evaporated to obtain a fresh crop.The several crops of crystals are united and dried at 108' C. They consist of about 70 per cent. sodium selenate and 30 per cent. sodium sulphate. The first crop contains more sulphate and a certain degree of frac-tionation is possible but is not worth while as sodium sulphate is not an objection-able impurity and any attempt to eliminate it reduces the recovery of selenium, which in no case exceeds 60 per cent. The fusion only occupies a few minutes. G. C. J. Determination of the Specific Gravity of Mineral Powders. M. Billy. { Compt. rend. 1913 156 1065-1067.)-To prevent the inaccuracies caused by the air retained between the particles of a mineral powder the water used in the deter-mination of the sp.gr. may be replaced by a solution of potassium hydroxide (of about Any bubble of gas will then no longer in-crease the volume of the powder but will be absorbed by the alkali solution. A special form of sp. gr. bottle for this purpose is described in which air is replaced by carbon dioxide at the same temperature and pressure. The sp. gr. of the alkaline solution is taken at 4" C. under the same conditions as that of the powder. strength) and the air by carbon dioxide. C. A. M INORGANIC ANALYSIS 297 New Proeess for Titration of Sulphurous Acid and of Sulphites in Presence of Thiosulphates. E. Bosshard and W. Grob. (Chem. Zeit. 1913, 37 465-466.)-When excess of mercuric chloride is added to a solution of a bisulphite, the solution previously neutral to methyl orange reacts acid towards that indicator, and continues to do so until sufficient alkali is added to convert the whole of the bisulphite into normal sulphite.The end-point is much sharper than that obtained in the usual titration with phenolphthaleln as indicator and the results are exact if Some sodium chloride is added to counteract the tendency of the mercuric chloride to become partially hydrolysed and react acid. In mixtures of normal and acid sulphite the former is first titrated with hydrochloric acid using methyl orange as indicator and the resulting solution of bisulphite is retitrated with caustic alkali, after addition of 1 grm. of sodium chloride and 300 C.C. of a saturated solution of mercuric chloride. The above observation of the effect of a large excess of mercuric chloride on the behaviour of bisulphites towards indicators shows that Feld’s method for the estimation of thiosulphates (Zeitsch.angew. Chem. 1911 24 290 ; ANALYST, 1911 36 435) would fail in accuracy in presence of sulphites or bisulphites. That method depends on the decomposition of thiosulphates in accordance with the equation Na,S20 + 2HgC1 + H,O = Na,SO + Hg,CI + 2HCl+ S. Mixtures of normil and acid sulphite with thiosulphate may be analysed as follows The normal sulphite is titrated with hydrochloric acid using methyl orange as indicator (l) and the resulting solution of bisulphite is re-titrated with caustic alkali in presence of phenol-phthalein (2). A further portion of the solution is then treated with a large excess of mercuric chloride and 1 grm.of sodium chloride and after addition of 30 C.C. of 20 per cent. ammonium chloride (to prevent precipitation of mercuric oxide) is titrated with caustic alkali using methyl orange as indicator. From the m n ~ n t of alkali used that due to the bisulphite originally present [calculated from the difference between titrations (2) and (l)] is deducted and the remainder calculated to thiosulphate in accordance with Feld’s equation. G. C. J. Detection of Tin. L. J. Curtman and M. Mosher. ( J . Amer. Chem. Soc., 1913 35 357-365.)-A study of the limits of sensitiveness of the mercuric chloride and ammonium molybdate tests and of the conditions which determine the maximum sensitiveness starting with a stannic solution containing antimony such as is separated in the course of a qualitative analysis.Under the most favourable con-ditions the mercuric chloride test will detect 0.2 mgrm. of tin in 5 C.C. The most favourable conditions are reduction of the tin in hydrochloric acid solution by heating 5 C.C. with one 14-inch cut iron nail for three minutes on the water-bath, and filtering into a tube containing 2 C.C. of saturated mercuric chloride solution. As much as 250 mgrms. of antimony are without influence on the result. The molybdate test is even more sensitive than the above test if carried out exactly as follows Five C.C. of the stannic solution which should contain about 1.5 C.C. of concentrated hydrochloric acid are heated on the water-bath with not more than 0.5 grm. of granulated zinc. When action has practically ceased the precipitated tin is dissolved by addition of a further 1 C.C.of hydrochloric acid and the heating continued. The solution is then diluted to lOc.c. and filtered into 1 C.C 298 ABSTRACTS OF CHEMICAL PAPERS of saturated ammonium molybdate solution. Under these conditions and in absence of antimony as little as 0.01 mgrm. of tin will give a faint but distinct blue colora-tion. The presence of 250 mgrms. of antimony reduces the limit of sensitiveness of the test to 0.05 mgrm. tin ; with 10 mgrms. of antimony present 0.03 mgrm. tin can be detected whilst 5 mgrms. of antimony are without measurable effect. The sensitiveness of this test is diminished by a higher concentration of acid than that above given. For example if the final solution contains 5 C.C.of concentrated acid in 10 c.c. even as much as 1 mgrm. of tin may be overlooked. G. C. J. Determination of Zine in Ores. D. J. Demorest. ( J . I n d . and Eng. Chem., 1913,5 302-304.)-The following modification of Low's method is applicable to any ore except Gahnite (zinc spinel) which must first be fused with sodium carbonate and borax glass. One grm. of the ore is heated for several minutes with 5 C.C. of hydrochloric acid 20 C.C. of nitric acid are added and heating is continued with the cover on until brown fumes cease to be evolved. One grm. of potassium chlorate is then added and the mixture is evaporated to dryness. About 50 c.c of wafer are added together with not more than 0-5 grm. of potassium hydroxide and the cake is broken up. The purpose of the potassium hydroxide is to neutralise any trace of nitric acid which otherwise would subsequently form ammonium nitrate in which cadmium carbonate is somewhat soluble.About 6 grms. of ammonium car-bonate are added and the solution is kept near to the boiling-point for several minutes. The precipitate which contains all the iron aluminium manganese lead and cadmium is filtered off? preferably by suction and washed with hot 5 per cent. ammonium carbonate. I t is then washed back into the beaker dissolved in 3 C.C. of hydrochloric acid a little potassium nitrite is added to reduce man-ganese dioxide and the solution is neutralised by means of caustic potash of which no excess should be used. Ammonium carbonate (5 grms.) is again added and the mixture heated to boiling for several minutes and filtered again through the original filter. Very small precipitates need not be redissolved and reprecipitated. The filtrate which should not exceed 200 c.c. is neutralised with hydrochloric acid of which an excess of 20 C.C. (sp. gr. 1-19> is then added. After heating to 70" C. hydrogen sulphide is passed through the solution which precipitates most of the copper without any co-precipitation of zinc owing to the high concentra-tion of acid. When no further precipitation occurs 20 C.C. of dilute (1 1) ammonia, (nearly but not quite enough to neutralise the acid present) are added slowly with constant stirring and the stream of hydrogen sulphide is continued for several minutes. This completes the precipitation of the copper Then without filtering or expelling hydrogen sulphide the solution is heated again to 70" C. and titrated with. ferro-cyanide using uranyl nitrate as indicator. Five tests of an ore issued by the U.S. Bureau of Standards as containing 31.41 per cent. of zinc gave results varying from 31.30 to 31.45 per cent. Addition of salts of cadmium copper and lead to the original solution had no influence on the results. 0. C. J

ISSN:0003-2654    

DOI:10.1039/AN9133800283

出版商:RSC    

年代:1913

数据来源: RSC

摘要:

298 ABSTRACTS OF CHEMICAL PAPERS ERRATUK-On page 81 of the Journal (February, 1913), line 17, f o r D. J. Daven- port, read D. J. Demorest.

ISSN:0003-2654    

DOI:10.1039/AN9133800298

出版商:RSC    

年代:1913

数据来源: RSC