摘要:

DECEMBER, 1912. Vol. XXXVII., No, 441. THE ANALYST. PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS. THE DETECTION AND ESTIMATION OF ARACHIS OIL. BY NORMAN EVERS, B.Sc., A.I.C. (Read at the Neeting, November 6, 1912.) DISCUSSION. THE PRESIDENT said that this process would be found to be a great advance upon Renard’s process in point of rapidity. I n order, however, to ensure accuracy, it was essential that all the modifications mentioned by Mr.Evers should be adhered to. Mr. R. Ross asked if it would be of any use to ascertain the refractive index of the residues obtained after cooling. He (Mr. Ross) had found this to be useful to a certain extent with butter containing coconut oil, though in that case its utility was limited owing to the variations in composition to which butter was liable.I n the case of olive oil, which appeared to be more constant in composition, the refractive index might be more helpful. Mr. P. A. E. RICHARDS asked how this process compared with Renard’s, more particularly with regard to the time required for a determination. Mr. L. M. NASH asked for a definition of “ saponified ” olive oil.Mr. RICHMOND asked whether the factors for calculating the quantity of arachis oil were arrived at from the samples of arachis oil referred to in the paper or from a more extended series of observations. I t seemed to him that in order to obtain accurate results the melting-points would have to be very carefully determined, and fully corrected, because, while the variation of the factor from 17 to 22 was appre- ciable, the difference of 2 degrees between the corresponding melting-points was only just outside the limits of experimental error.Mr. W. B. SAVILLE said that he had used Bellier’s process, and had found it to work very well, though when the proportion of arachis oil was considerable the results were liable to be low. He had usually found the melting-point of the538 ARCHBUTT: AX APPARATUS FOR TESTING WATER fatty acids to be 73’ C.He had tested almond, apricot, cotton-seed, maize, peach- kernel, castor, and sesame oils by Bellier’s qualitative method, and in each case found a perfectly dear solution on standing for one to two hours at 17’ to 19’ C. Rape-seed oil invariably gave a turbidity due no doubt to the small percentage of arachidic acid present in the oil.Arachis oil could be determined with reasonable accuracy by the author’s process when in admixture with these oils. Mr. EVERS, in reply, said that he had not determined the refractive index of the crystals, but should think it quite probable that it would afford valuable information. The time required for a determination of arachis oil by this process was merely a few hours, and little attention was required, whereas the Renard process took considerably longer.“ Saponified ” olive oil consisted of the oil extracted from the marc after the larger proportion of the oil had been expressed. I t contained a good deal of free fatty acid, and in order to remove this it wag washed with alkali, which operation gave rise to the term “saponified.” The factors were arrived at from the results yielded by the samples of arachis oil referred to in the paper.The meltingpoints had not been corrected for barometric pressure or the height of the column of mercury above the liquid, and this possibly might account for the differ- ence between his figures and Mr. Saville’s. In Bellier’s process he had never found the melting-point (uncorrected) to be higher than 71’ C . In his experience the melting-point could be determined accurately within half a degree. There was a, good deal of difference in the proportion of fatty acids obtained when the melting- point was 71’ C. and when it was ‘72’ C., and this accounted for the difference in the factor.

ISSN:0003-2654    

DOI:10.1039/AN9123700537

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

538 ARCHBUTT: AX APPARATUS FOR TESTING WATER AN APPARATUS FOR TESTING WATER BY MEASUREMENT OF ITS ELECTRICAL CONDUCTIVITY. BY L. ARCHBUTT, F.I.C. (Read at the Meeting, November 6, 1912.) AN apparatus known as the Digby and Biggs ‘‘ Dionic ” Water-Tester, designed primarily for detecting leakage in surface condensers, was recently brought to my notice; and having proved of some interest, I thought it might also interest some members of the Society.The usefulness of the apparatus depends upon the well- known fact that the conductivity of pure water is extremely small, and that the presence of any salt, acid, or base, in solution at once increases the conductivity. In very dilute solutions Kohlrausch found that the conductivity due to a given substance increases in proportion to the amount dissolved.For stronger solutions, however, this is not correct. Different salts also have different conductivities, The value of the apparatus depends upon its extreme sensitiveness in detecting mere traoeg of dissolved electrolytes, and in the rapidity with which a test can be made.BY MEASUREMENT OF ITS ELECTRICAL CONDUCTIVITY 539 Referring to the diagram, the water to be tested is poured into the funnel F, and rises into the two arms of the U tube to the level fixed by the overflows 00.The electrodes A and B are short tubes of platinum-foil, and the current is supplied by the hand-driven dynamo E, fitted with a conatant-speed friction clutch, which maintains a constant E.M.F. of 100 volts, within volt, when the handle W is turned at a rate faster than that at which the clutch is designed to slip.The current passing through the water is measured by the direct-reading conductivity meter M. Temperature is measured by a thermometer immersed in the water in the funnel. To make a test, all one has to do is to empty the U tube and funnel by opening the clip D, flush the apparatus with the water to be tested, then close D and continue pouring water into the funnel until it overflows at 00.Then turn the P handle of the dynamo until the clutoh is slipping, and at once read off the maximum deflection indicated by the needle. The current generated is very small, not exceeding 0.004 ampere under normal conditions. The back E.M.F. due to polarisation of the electrodes is allowed for in the calibration of the meter, the scale units of which, ranging from zero to 2,000, represent approximately in value the reciprocals of megohms.The standard tempera- ture for testing is 20" C., and a scale of corrections supplied with the instrument saves the trouble of adjusting the temperature of the water so long as it is between loo and 40" C. I t is therefore necessary only to take the temperature of the water and correct the meter readings by the corresponding factor.The usefulness of the apparatus in a laboratory or works, and also its limits, will be apparent by the results of the following tests made in my laboratory :540 ARCHBUTT: AN APPARATUS FOR TESTIKG WATER Conductivity (corrected to 20" 0.)) Distilled Water. Unboiled distilled water ...... ... ... ... 2.7 Another sample ... ... ... ... ... ... 6.0 Another sample, after boiling to expel dissolved gases and cool- ing in air ... ... ... ... ... 1.5 Several samples thoroughly boiled in a silvered flask and cooled Unboiled water after passing GO,-free air through for about Saturated with GO,, and'excess oigas shaken out at 20" C. ... The above figures show that the ordinary distilled water of the faboratory, if boiled to expel GO,, has a conductivity in this apparatus of 1.0 to 1.5.Derby Town water, with a soap hardness of 18*4O, total solids about 25 grains per gallon, chlorine about 1.15 grains, and SO, about 4 to 5 grains per gallon, had a, conductivityof 552. 3 C.C. of this hard water diluted to 500 C.C. with boiled distilled water, and contain- ing about 0-15 grain of total solids per gallon, had a, conductivity of 4-2.This wabr gave a faint opalescence with silver nitrate in the cold, which disappeared on heating, and no perceptible precipitate with barium chloride, from which it is evident that a. slight leakage of the hard water into the distilled water through a crack in the tin condensing worm, as sometimes occurs, would be detected by the conductivity meter rather more readily than by the silver nitrate test, and the test would be quantitative. ...... ... *. . ... 1.0 to 1.5 in vacuo ... one hour ... ... ... 1.0 44.0 In the next table I give the conductivity of a, few public water supplies. Glasgow, from Loch Katrine ... ... ... ... Manchester, from Thirlmere ..- ... ,.. ...Blackburn ... ... ... Birmingham, from Rhayader ... ... ... ... Dewsbury ... ... ... ... East Surrey water, from wells in &e chalk, softened by Clark's process ... ... ... ... ... Derby: mixture of spring water and Derwent Valley water ... ... ... West Middlesex water (Meiropolitan Water Board) ... ... (hardness, 11.9 degrees) ... ... ... ... Derby : spring water only (hardness, 184 degrees) ...... Conductivity (20" C.). 34 48 65 80 109 135 390 380 552 The above, except the Derby samples, are from a pamphlet published by Mesas- Evershed and Vignoles, Ltd., the makers of the apparatus. As there is a tendency for an apparatus of this kind to be offered 818 a substitute for chemical analysis, I have prepared a few soIutions and tested their conductivity.The saline solutions (except the MgSO,) were made with the salts usually met with in concentrated boiler waters. Being made of equal strength, they show how the conductivity varies with the different salts, also that with solutions of the strengths stated the conductivity is not proportional to the amount dissolved (compare N a2 S 0,). The acids and bases show enormous variations; the first three, also the caustic soda solution, were made up so as to have about the same conductivity as a solution of common salt containing 50 grains per gallon; the acetic acid and ammonia wereBY MEASUREMENT OF ITS ELECTRICAL CONDUCTIVITY 541 NaOH, 3.2 C.C.N. soda diluted to 500 C.C. ... ... ... NH,, 35.45 C.C. N. ammonia diluted to 500 C.C. random mixtures. grains per gallon for eonvenience of comparison.conductivity of the water tested may be influenced by a slight acidity or alkalinity. I have stated all the strengths in normal " terms as well as in The figures show how greatly the 17.9 84.7 (174.5 NH40H) Some Saline Solutions. Total Solids, 130' C. Grains of Anhydrous Salt per Gallon. Non-Volatile Solids. Chlorine a8 Chloride. Conductivity (20" C.) .NaCl ... ... ... ... Na,CO, ... ... *.. Na2S04 ... ... ... w 3 0 4 MgSO4 ... ... ... ... ... ... ... ... ... NaN03 50.0 50.0 50.0 100.0 50.0 50.0 Acids. IIC1 1.6 C.C. normal acid diluted H2s04) HNO, to 500 C.C. ... ... CH,CbOH, 26.5 C.C. N. acid diluted to 500 C.C. ... ... ... I 8.16 10.98 14-37 222.60 Bases. Conductivity (20" C.). 1,287 1,236 1,020 1,751 900 954 1,260 1,127 1,255 400 1,214 325 Lastly, non-electrolytes, such as cane-sugar, might be present in water in large quantity, and would entirely escape detection.Within limits, however, and applied to specific purposes, the apparatus is extremely useful, as is shown by the following tests : CONCENTRATED WATER FROM STIRLING BOILER. Grains per Gallon. 19.8 51.9 53.7 60.1 69-1 73.3 75.6 112.4 Grains per Gallon.19.0 47.9 52.1 54.2 64.0 67.2 72.2 108.8 I Grains per Gallon. 3.6 14.3 16.7 17.5 11 0 7 16-0 18.8 24.2 I 445 1,113 1,166 1,261 1,272 1,431 1,579 1,961 I542 ARCHBUTT: AN APPARATUS FOR TESTING WATER The above samples were taken on various dates, and do not represent any particular series. They are arranged in order of conductivity. The water used in the boiler was partly condensed water and partly a hard water used 8s ‘( make-up.” The conductivity, though not strictly proportional to the non-volatile solids, is fairly proportional, whilst the chlorine is not. Suppose it were desired, for example, to keep the concentration of the water in this particular boiler below 100 grains per gallon in order to prevent “priming,” it is evident from the above figures that the conductivity test is a better, simpler, and more rapid indicator of the amount of saline matter in solution than is the chlorine, which is about the simplest analytical test that could be made.I t is obvious that this apparatus may be extremely useful for detecting leakage of surface condensers. Condensed water from the hot wells of condensing engines will have a quite low conductivity, if there is no leakage, say 3 to 5.Cooling water will generally be hard, and will have a much higher conductivity. This applies especially to sea-water, the conductivity of which is about 50,000 on the scale of this instru- ment. Having ascertained, therefore, the conductivity of the cooling water used, and of mixtures of this with distilled water, a curve can be constructed from which the percentage of leakage water corresponding to any observed conductivity can be read off.DISCUSSION. Mr. F. EVERSHED asked whether the President had found this apparatus to be useful for controlling the addition of chemicals in softening water-one of the purposes for which the makers claimed that it might be used with advantage in place of the soap test.He (Mr. Evershed) should think that in simple cases, where the addition of lime only was required, the apparatus could be so used; but in more complicated cases, when soda, with or without lime, was used, the conductivity would not necessarily be diminished in the final result, though even here such method of testing might be found useful, if applied with discrimination. In sugar- anrtlysis a conductivity test would often measure inorganic impurities with sufficient accuracy and far more quickly than an ash determination. This last idea, was not new, a paper on the subject having been read by Mr.Hugh Main at the International Congress of Applied Chemistry in 1909, but Mr. Main used the Kohlrausch apparatus, which required a supply of electricity from an outside source, while the apparatus now described had the advantage of being self-contained, the current being produced within the box by merely turning a handle.Various other applications of this instrument would no doubt suggest themselves. Mr. RAYMOND Ross said that in Lancashire there were many streams which, in addition to their ordinary main flow, received colliery pumpings, and also varying amounts of peaty water from the surrounding hills.Such water often had to be used for boiler purposes, and the variations in its oomposition presented a problem in regard to which the apparatus referred to might perhaps be useful. Mr. G. N. HUNTLY said that the principle on which this apparatus was based had already been applied in practice for Borne time.There is a well-known oil- removing plant, in which the removal of oil from condensed water is effected byBY MEASURENENT OF ITS ELECTRICAL CONDUCTIVITY 543 passing a continuous current between iron plates, tbe oil being carried down by the oxide of iron that was thus formed. It is found, incidentally, that if a fuse is placed in the circuit, any leakage of cooling water into the condenser is infallibly indicated by the blowing out of the fuse.The difference in conductivity w a ~ so great that a heavy fuse is readily blown out. Mr. W. T. BURGESS said that the conductivity measurements were, of course, mainly influenced by the mineral constituents, and gave no indication of the organic purity, or otherwise, of the samples tested.Mr. J. H. B. JENKINS said that the apparatus used by Mr. Main answered very well for the estimation of ash in sugar, the determinations being made very quickly. It was quite easy to experiment in this direction without any special apparatus beyond that which was ordinarily a t hand in the laboratory. If the terminals of an induction-coil were connected with a Wheatstone bridge-the galvanometer being replaced by a telephone receiver-and if into the resistance space of the bridge were introduced a platinum dish with a small electrode kept about & inch away from the bottom of the dish, samples of water could be introduced into the dish one after the other and tested by noting when the sound in the telephone receiver disappeared. In this case, of course, it was the resistance and not the conductivity of the water that was measured. The PRESIDENT, in reply, said that he could quite understand that there might be circumstances in wbich this apparatus would be useful in connection with water- softening, but, on the other hand, conditions could easily be imagined in which it would be practically useless, Probably in a case such as that mentioned by Mr. ROSS, if the conductivities of the different sources of water were known, it would be possible to determine approximately the extent to which the various kinds of water were present, but here again there would be limitations.

ISSN:0003-2654    

DOI:10.1039/AN9123700538

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

BY MEASURENENT OF ITS ELECTRICAL CONDUCTIVITY 543 THE EXAMINATION OF CHINESE AND JAPANESE WOOD OIL. BY A. CHASTON CHAPMAN, F.I.C. (Read at the Meeting, November 6, 1912.) CHINESE WOOD OIL. DURING recent years the trade in Chinese wood oil has assumed considerable proportions, and its chemical examination has consequently become a matter of importance both to buyers and sellers. This fact, coupled with the very in- teresting nature of the oil itself, has induced a number of chemists to devote themselves to its investigation, and a number of papers dealing with the general chemistry and analytical examination of the oil have been published in English, American, and Continental journals.Notwithstanding this, it cannot be said that chemists are in complete agreement as to the properties which should chsracterise the genuine oil, and for Borne time past trade disputes culminating in544 CHAPMAN : THE EXAMINATION OF arbitration cases have been of somewhat frequent occurrence.Having for several years been much interested in the examination of this oil, I have thought that it might not be without interest if I were to place some of my results and conclusions before this Society, This oil is obtained, as is well known, from the seeds of the Aleurites cordata -a tree whichgrows extensively in China and other parts of the East.Other species of Aleurites-as, for example, Aleurites Fordii-also occur, and it is probable that the commercial oil often represents a product obtained from more than one species. I shall revert to this later.According to a recently-published United States Consular Report, it would appear that Szechuan oil is preferred in Hankow on account of its lighter colour, and that the province of Szechuan accounts for 30 per cent. of the total production, Hunan and Kweichow producing 50 per cent., and Hupei (chiefly in the region of Ichang) 20 per cent. Lewkowitsch (“ Chemical Technology and Analysis of Oils, Fats, and Waxes,” fourth edition, vol.ii., p. 60) states that, in the preparation of the oil, the seeds are roasted in a flat dish over a naked fire, are then crushed between stones, and the oil expressed from the crushed mass in primitive wooden presses. This account of the preparation of the oil is confirmed by several independent statements made to me by persons who have resided for a considerable time in China, and who are familiar with the method of expression adopted by the natives.This crude method of obtaining the oil, considered in conjunction with the fact that the production is in the hands of a large number of very small producers, and that the oil in question is a powerful drying oil, is of considerable importance, since it affords a partial explanation of the somewhat widely varying analytical results which have been obtained by chemists in the examination of this oil, and which have been recorded in the literature of the subject.The small quantities of oil produced by the various individuals are collected by Chinese middlemen, and sold to merchants in Hankow, who clarify the oil in tanks, and then pack it for shipment.In colour the oil varies very considerably, the dark oils having as a very general rule a more markedly unpleasant taste and smell than the paler ones. The analytical determinations which are of the greatest importance are the specific gravity, the iodine value, the refractive index, the viscosity, and the manner in which the oil polymerises when heated.I propose to refer in the first place to these estimations, and to deal later in the paper with several supplementary methods of examination. Speci$c Gravity.-Since the specific gravity of Chinese wood oil is higher than that of any other fatty oil, with the exception of castor oil and possibly of tallow- seed oil, the determination of that number is clearly of great importance.The numbers recorded in literature vary somewhat considerably, but many of the earlier ones are of less value than they might otherwise have been owing to some apparent confusion between the Chinese and the Japanese oil, and also to the fact that in some cases the nature of the ratio oil : water is not clearly stated. Among later and more reliable numbers may be mentioned those of Lewkowitsch (“Chemical Technology and Analysis of Oils, Fats, and Waxes,” fourth edition, vol.ii., p. 63) 0.9412 to 09418; Kreikenbaum (J. Ind. and Eng. Chem., 1910, 2, 205), 0.9401 toCHINESE AND JAPANESE WOOD OIL 545 09454; and F. Browne (Chem. News, 1912, 106, 15), 0.9402 to 0.9431. Jenkins (ANALYST, 1898, 23, 113), in a paper communicated to this Society, gives 0.9343 and 0.9385 for two samples ; but as he uses the terms 6 c Japanese wood oil,” I‘ Chinese wood oil,” and ‘‘ tung oil,” as synonymous, the value of these numbers is considerably diminished.Rowland Williams (J. Soc. Chem. Ind., 1898, 17, 304), in a paper entitled ‘‘Japanese Wood Oil,” gives numbers ranging from 0-9413 to 0.9432, and it is tolerably clear from these and the other results he gives that these were samples of the Chinese oil.Boughton (Sewenth Int. Cong. App. Chem., 1909, Section 1, 89) records numbers for five samples of China wood oil which he had received directly from the United States Consul at Hankow, and for two further samples which had been pressed for him from the nuts. Six of these gave numbers ranging from 0.9408 to 0.9428 at 15-6” C., but one of the five samples from Hankow had at the same temperature a specific gravity of 0-9346, which is far below any number recorded by other observers for the Chinese oil, and which must tend to throw considerable doubt on the purity of the specimen.Iodine ‘VaZue.-Sinoe this number depends so largely on the precise method adopted and upon the conditions under which the test is carried out, and as this information is not given in many of the earlier papers, it would serve no useful purpose to refer in detail to the numbers recorded, but it may be said that numbers varying from 150 to 170 have been given.In some cases, however, it is probable that the Japanese oil had been used. Boughton (Zoc. cit.) has made a special study of the extent to which the iodine value varies with different experimental conditions, and finds that variations in temperature have much less effect on the Hub1 than on the HanGs method.For several reasons I prefer the Wijs method, and the iodine values recorded in the table given below have been determined in the following manner :-About 0.1 grm. of the oil was taken, dissolved in 20 C.C. of purified oarbon tetrachloride, 30 C.C.of the ordinary Wijs solution were added, and the reaction allowed to proceed for three hours in the dark. Refractive Index.-Since the refractive index of wood oil is considerably higher than that of any other fatty oil, the determination of that value is of the greatest importance from the analytical point of view. Comparatively few numbers are recorded in the literature of the subject, but it may be mentioned that Boughton, gives 1.5238 at 15.6” C., which is approximately equal to 1-521 at 20’ C., the tem- perature at which my own observations have been made.Viscosity (Time of E$Zuz).-As the viscosity of tung oil is considerably higher than that of any other fatty oil which would be likely to be used as an adulterant, a.determination of that number is very useful, although, as will be seen from the results recorded below, there are wide variations even in the case of oils of undoubted genuineness. I t must not be overlooked, moreover, that the viscosity of wood oil may be very considerably increased by heating it to a temperature short of that required for its solidification. The viscosities given in the following table have been.determined in a standard Redwood viscosimeter at 15.5” C. From a circular recently issued by the New York Produce Exchange, it would appear that C. V. Bacon has made a large number of viscosity determinations, using a Tagliabue instrument, and that he attaches considerable importance to the results.546 CHAPMAN : THE EXAMINATION OF Heat Polymerisation Test.-One of the most interesting and characteristic prop- erties of wood oil is that of solidifying to a jelly when heated to a sufficiently high temperature, a change which appears to be due to the polymerisation of the glyceride of elsomargaric acid, of which the oil largely consists.Since the hardness of the resulting jelly is considerably diminished by the addition of other fatty oils, it will be seen that a test based upon this property is capable of yielding very important indi- cations.Several attempts have been made to make such a test quantitative. Thus, in a circular issued by the New York Produce Exchange, to which I have referred already, a method devised by C. V. Bacon is tentatively put forward as a method to be adopted in the United States.Since this test has not, so far as I am aware, been published in any scientific journal, it may be well to give the following description of it : I ‘ Into a test-tube of 3 inch diameter and 4 inches in length there are transferred about 10 C.O. of pure China wood oil; into another test-tube there is transferred a similar volume of pure China wood oil adulterated to the extent of 10 per cent.A sample of the oil to be tested is treated in a like manner, and these are placed in a proper support and immersed in an oil bath which has a temperature of about 288” C., so that when the tubes are in it a temperature of 280° or 285O C. (maximum) can be maintained. The oil bath containing the tubes is maintained at this tempera- ture for exactly nine minutes, the tubes are then withdrawn and the test sample is compared with the pure oil, and the same oil adulterated with 5 and 10 per cent.of foreign oil. After the tubes are withdrawn from the oil bath, each tube should be stabbed from the top to the bottom with a small bright spatula. Pure oil will give a hard, clean cut, and when the knife is withdrawn the incision will look like a straight line, while an oil having as low as 5 per cent.adulteration will invariably be softer, and the incision will have a peculiar feathered effect; an adulteration of 10 per cent. will be soft and pusby, while an adulteration exceeding 12 per cent. in many instances will remain entirely liquid.” In the same circular a further test is described as being one at present used by New York importers and varnish-makers.This is described in the following words : “For Hankow and Shanghai wood oil, 100 grms. should be heated in an open basin (6 inches in diameter) as soon as possible to a temperature between 540” and 560° F., and if it solidifies in about six to six and a half minutes, cuts dry, and is firm in body, without discoloration and without being sticky, it should be passed as a good delivery.For Canton and Hong-Kong wood oil deliveries, the time should be from four and a half to five and a half minutes in an open basin as above. Should a longer time be taken by presumably pure wood oil, other test8 confirming purity shall be positive.” This latter test is, in my opinion, an unsatisfactory one, since it is impossible to control the high temperatures necessary or to keep them at all constant, owing to the shallowness of the containing vessel and the large cooling surface.As soon, more- over, as golidification commences, it is impossible to continue the stirring, with the result that, if the application of heat be continued, the bottom portions of the oil will &&r, whilst the top portions are still in a soft and sticky condition. The main objection to Bacon’s method lies in the fact that it involves comparison with a,CHINESE AND JAPANESE WOOD OIL 547 standard sample, which in turn implies that wood oil is practically constant in its polymerising properties. If all genuine wood oils were known to polymerise in precisely the same manner, there would be some justification for drawing definite conclusions from the indications furnished by such a method, although even then they could not, from their very nature, be more than roughly quantitative.There can b4 no doubt, however, that such is not the case, and consequently it is not wise to attempt to draw too definite conclusions from the results of the heat test, valuable as its indications are.I t may be mentioned in this connection that F. Browne (Zoc. cit.) has described a method of applying the heat test which is, in my opinion, superior to that of Bacon, and less open to criticism, inasmuch as he recognises that genuine oils may vary somewhat in their setting properties ; but the temperature to which the oil is heated is so high that the differences between the setting period of a genuine oil and one adulterated with, say, 10 per cent.of some other fatty oil is only two or three minutes. As the result of a considerable amount of experience, I have been led to attach more importance to the hardness of the jelly obtained under standard conditions than to the time required for bringing about polymerisation, and 1 have devised the following method of carrying out the test, which is capable of yielding definite and concordant results : About 5 C.C.of the oil to be examined are introduced into each of two test-tubes 6 inches long by Q inch diameter. These are then immersed in a bath containing melted paraffin wax at a temperature of approximately looo C. The temperature of the bath is then raised to 250" C., taking about fifteen minutes for the operation.As soon as that temperature is reached the time is noted, and the source of heat adjusted so that the temperature of the bath is maintained constant at 250" C. At the end of half an hour one of the tubes is withdrawn, allowed to cool, and, when cold, is broken, and the jelly examined. The other tube is kept in the bath at 250' C. for a further period of half an hour, at the end of which time it also is with- drawn and allowed to cool ; it is then broken, and the hardness of the jelly observed. Chinese wood oil of good quality should give at the end of half an hour a fairly firm jelly, which, at the end of one hour, should become quite hard.I t is advisable in all cases to carry out comparison tests alongside of the oil under examination, using for the purpose a sample of oil known to be of good quality. I have not found it possible to express the hardness of the solidified cylinders by means of numbers, but with a little experience it is very easy to distinguish between a sample of genuine oil and the mme oil containing a small percentage of some fatty oil, such as soya bean or sesame.In referring to the polymerisation experiments, I have used the words L L very hard," "hard," and " fairly hard," to denote the con- sistency of the polymerised oil, since such expressions are quite sufficient for the purpose.In addition to the degree of hardness of the solid cylinders of oil, some attention should be given to their physical characters. When cut with a knife or broken across, the cut or fractured surface should be smooth and free from stickiness, and small portions when rubbed in the hand should break down completely into a, soft crumbly mass, which should not adhere to the fingers.In the following table I give the results of the examination of seventeen samples of Chinese wood oil from Hankow :548 CHAPMAN : THE EXAMINATION OF Sample.Average ... Iodine Value. 169.9 168.4 166.5 166.4 168.8 170.0 168.6 171.0 169.7 173.0 176.2 172.6 174.2 173.7 172-8 169.5 169.6 170.6 Sp. Gr. 1 S"/l 5" C. 0.9419 0.9406 0.9426 0.9417 0.9430 0.9440 0.9416 0.9414 0.9437 0.9420 0.9417 0.9429 0.9427 0.9430 0.9440 0.9420 0,9433 0.9425 Saponification Value. 196.6 193.8 194.3 193.0 195.6 194.5 193.0 192.0 194.1 192.5 192.0 196.0 194.6 195.0 194.6 195.2 195.2 194.2 Refractive Index a t 20" c.1.5207 1.5181 1.5190 1-5170 1.5195 1.5180 1.5150 1.5170 1.5176 1.5165 1.5168 1.5180 1.5182 1.5194 1.5193 1.5160 15187 1.5179 Time of Efflux a t 15.5" C. Seconds. 2178 1636 1946 1880 2017 1849 - - 1997 1722 1605 1740 1690 1820 2047 1804 1820 1850 Polymerisation. One Hour at 250" C. Very hard. Hard. Fairly hard.. Verykwd.Hard. Fairly hard.. Hard. Y 9 9 ) 9 9 9 , 3 , 9 9 9 9 Veri'hard. On reference to the above table it will be seen that the specific gravities of the seventeen oils in question range from 0.9406 to 0.9440, and these results, which are confirmed by a very much larger number, may be regarded as approximately the limits of variation for the genuine oil. I am not prepared to say that no pure Chinese wood oil has a specific gravity below the lower limit mentioned above, but a result appreciably below 0.940 is in itself sufficient to cause a sample of oil to be regarded with suspicion.With regard to the iodine value, it will be seen that the average of the results is 170.6, and it may be taken that a genuine oil should not give numbers differing from that by more than a few unitB, using the Wijs method, and adopting the conditions.I have specified. The saponification value is not, in the majority of cases, of very much import- ance; but as it may sometimes be useful, it should always be determined. I t is interesting to note that my limits agree fairly closely with those recorded by Rowland Williams (190.7 to 196.1)) in the paper to which reference has already been made.The refractive index rarely falls below 1.517, and an oil giving a result below 1.515 at 20" C. should be regarded with suspicion. As I have already pointed out, the viscosity of this oil appears to vary within glomewhat wide limits, but the numbers I have given will suffice to indioate the kind of result which may be expected in the case of genuine oils.A paper dealing with the oil derived from the seeds of Aleurites Fordii having. appeared in the Bulletin of the Imperial Institute (1907, 5, 134), I wrote to ProfessorCHINESE AND JAPANESE WOOD OIL 549 W. R. Dunstan asking him if he could let me have some of this oil. Although he had none of the oil left, he kindly supplied me with a quantity of the nuts of Aleurites Fordii, as well as with some of the Aletirites cordata, and I desire to take this opportunity of expressing to him my thanks for this courtesy.From each of these I prepared a quantity of oil by petroleum extraction, and found that the former had a specific gravity of 0.9427, and the latter of 0.940. Both these oils, however, were abnormal in respect of their polyrnerising properties, and for that reason I have not included the other analytical results, Whether this abnormality is due to the method of extraction adopted or to the fact that the nuts were several years old, I am not prepared to say, but it is certain that both oils were markedly inferior in their polymerising power to any genuine sample of the commercial oil that I have examined.This result gives point to the remark I made in criticising Bacon's method, that all samples of genuine wood oil cannot be expected to polymerise in exactly the same manner.There is no doubt that Chinese wood oil, like all other oils, varies within certain limits, both in respect of its chemical and physical pro- perties. As a matter of fact, botanical and geographical origin, climatic variations, differences in treatment during the manufacture of the oil, and the duration and conditions of storage, unquestionably affect its properties, and I am, therefore, very strongly of opinion that it is not advisable in determining its purity to adopt any single method which involves a comparison with an oil possessing certain properties in a maximum degree.Of course, if any purchaser desires to buy wood oil on the results of the polymerising test alone, I see no reason why he should not do so.It must not be overlooked, however, that it may quite well happen that oil which does not reach a certain high polymerisation standard may nevertheless be free from adulteration. Sup p lementary Tests . Drying Test.-About 0.3 to 0.4 grm. of the oil is spread in a thin fiIm on a glass plate and heated in an air-bath for one hour at a temperature of 105" C.The genuine oil dries completely to a crinkly, rubber-like substance, which can be removed without difficulty from the glass surface, and which is free from softness or oiliness. Under these conditions the oil gains from 2 to 3 per cent. in weight. Iodine Test.-As is well known, if a solution of iodine in chloroform or some other suitable solvent be added to the oil, almost immediate solidification takes place.This is a very useful qualitative test, and McIlhiney (J. Ind. and Eng. Chem., 1912,4, 497; ANALYST, 1912, 410) has recently been carrying out experiments with the object of ascertaining whether it might not be made quantitative. I n the method he has put forward tentatively, an excess of a solution of iodine in glacial acetict acid is added to a weighed quantity of the oil dissolved in the same solvent.The products of the reaction are then extracted with petroleum ether, and the amount of oil which has not undergone solidification, and which is soluble in the petroleum, is weighed. I have made a few experiments with this process, and have obtained results which appear to indicate that when this test is fully worked out it will be found to be a very useful one. Optical Activity.-A number of samples of oil which I have examined have been practically inactive, although in one case a slight dextro rotation (+ 1' for a,550 CHAPMAN : THE EXAMINATION OF 100 mm.tube and sodium light) was observed. Tallow-seed oil, on the other hand, is known to be appreciably laevo-rotatory. Bromine Thermal Value.-This number was determined by the method recom- mended by Hehner and Mitchell-viz., by adding 1 C.C.of bromine to e solution of 1 grm. of the oil in 10 C.C. of chloroform, the mixture being kept in a vacuum jacketed test-tube. The following results were obtained in the case of six samples of the oil : Rise in Temperature."C. Sample A ... ... ... 24.10 Sample B ... ... ... 24.76 Sample C ... ... ... 25-57 Sample D ... ... ... 24-98 Sample E ... ... ... 24-72 Sample F ... ... ... 24-56 Iodine Value. Rise in Temperature. ... 7.00 ... 6-80 . a . 6-51 ... 6.92 ... 7-00 ... 7.17 For purposes of comparison I give below the values obtained in the case of three samples of the Japanese oil : Iodine Value.Rise in Temperature. Rise in Temperature. "C . Sample A ... C . . ... 24-70 ... 6.40 Sample B ... ... ... 23.10 ... 6-45 Sample C (prepared in laboratory) 24-50 ... 6.27 I t will be seen that the average value (6.9) obtained by dividing the iodine value by the rise in temperature is substantially the same as that given by Jenkins (Zoc. cit.) in his paper on wood oil.There appears, however, to be no reasonable doubt that the two samples of oil with which Jenkins worked were Japanese oil, and for this it will be seen that my average value is somewhat lower-viz., 6.37. In both cases, however, the value is appreciably higher than with the great majority of fatty oils. Hexabromide Test.-It has been shown independently by Hehner and Mitchell, and by Jenkins, that when wood oil is treated with bromine under the conditionhi laid down by the former authors (ANALYST, 1898, 23, 310), no ether-insoluble bromide is formed.I have confirmed the correctness of these observations, working with the seventeen samples of Chinese wood oil referred to above, and with four samples of the Japanese oil. I n no case was there the slightest formation of pre- cipitate, even when the flask containing the mixture was kept in iced water and allowed to stand for twenty-four hours.On the other hand, perilla, candle nut, and tallow-seed oils all yield more or less hexabromide ; and as these oils are possessed of high drying properties, and are produced in large quantities in China and the East, the test may be one of considerable importance, and is, in any case, one which ahould always be applied.I n addition to the above-mentioned oils, it would, of course, aid in the detection of linseed oil. Tee-seed oil, as might be expected, does not yield any insoluble bromide, I have been able to detect with considerable ease the presence of 10 per Gent. of perilla oil, but it should be mentioned that this oil yields a much larger proportion of the hexsbromide than any of the oils mentioned above, with the possible exception of linseed.The oil should, of course, be tested in dl cases for those fatty oils such 8s cottoa-CHINESE AND JAPANESE WOOD OIL 551 Sample. No. 1 ... No. 2 ... No. 3 ... Saponifica- Refractive Tiyu:f Polymerisation pzE $jl!??b. tion Valuc.Index at 15.50 c. Two Hours at 250" C. at 2oo '* Seconds. 158-0 0.9377 195.2 1.5083 1230 Soft . 149-0 0.9400 193.4 1-5052 1620 Soft. 151.8 0.9349 196-3 1.5034 - Very soft.552 CHAPMAN : THE EXAMINATION OF When heated for two hours at a temperature of 250” C., this sample had not solidified, but was still a viscous oil. On reference to the above numbers it will be seen that with the possible exception of No.2 ” the specific gravities are very appreciably below those of the Chinese oil, which confirms the statements of Lewko- witsch and other observers. The same applies to the iodine values and refractive indices. Even greater than these differences, however, are the differences in poly- merising properties, the Japanese oil remaining quite liquid under conditions such as suffice to convert the Chinese oil into a veryhard jelly, which crumbles when rubbed between the fingers and becomes a dry powder.My best thanks are due to my assistant, Mr. Frederick T. Harry, for his valuable help in connection with this work. DISCUSSION. Mr. HEHNER remarked that when the fatty acids from this oil were dissolved in acetone and cooled, magnificent crystals resulted.This property was possessed by no other fluid oil. The crystals retained their form as long as the vessel was kept closed : when it was opened they generally collapsed immediately into a jelly-like mass. Another remarkable feature of wood oil was that it was not capable of forming a hexabromide like other oils having a high iodine absorption-for example, linseed oil and fish oils-which might be used as adulterants.Probably the diflerences in the extent to which polymerisation took place were due to small quantities of botanical impurities which exercised a catalytic action. Mr. J. H. B. JENKINS said that when he examined these oils fifteen years ago the quantity of them on the market was very small, and consequently his investi- gation had been much more limited than that of Mr.Chapman. He should like to emphasise what Mr. Hehner had said as to the bromine addition value : the absence of any insoluble brominated compound markedly distinguished these oils from other drying oils. It was the indirect result of Mr. Hehner’s and Mr. Mitchell’s valuable work on the bromine reactions that had called his attention to the peculiarities of these oils.Following those authors’ method, he bad compared the iodine values and heat of bromination of a wide range of fatty oils. The usual factor for converting the one into the other was found, except in the case of a strange oil which was put into his hands by a friend as “Japanese Wood Oil.” In that case, however, the factor was distinctly abnormal, and led him to think that iodine must have some exceptiona,l action on the oil.He consequently tried the effect of adding a saturated solution of iodine to the oil dissolved in chloroform, and, whilst stirring, he found the whole was converted into a jelly. The other abnormal feature disclosed in his early examination of those oils was the very high refractive index. Both samples he had dealt with (ANALYST, 1898, 23, 113) were supplied as “Japanese Wood Oil ” ; but at that time, in common with many others, he had looked upon Japanese Wood Oil and Chinese Wood Oil as the same.Mr. E. R. BOLTON asked what was the lowest iodine value that Mr. Chapman had found in what might be accepted as genuine wood oil. I n making some experiments with a sample of Chinese wood oil which he believed to be genuine,CHINESE AND JAPANESE WOOD OIL 553 but of which the iodine value was very low, he had found that with this oil the time required for complete absorption was very short, about a quarter of an hour’s contact with Wijs’ solution being sufficient, so that some of the low figures that had been published could hardly be attributed to insufficient time being allowed for absorption.Mr. R. Ross remarked that the specific refraction (‘- - of the various oils referred to in the first table worked out st 542 to 550, and of those in the last table at 537 to 538, the figure for the oil expressed by Mr. Chapman himself being 534. He had found the McKinley process of bromine absorption to be very useful, since it showed both the addition products and the substitution products, and very often enabled one to account satisfactorily for certain differences that were observed.Mr. L. M. NASH said that at the present moment the iodine value by itself was of no assistance in judging wood oil, since linseed oil was about &4 per ton cheaper than wood oil. Mr. Chapman’s figures confirmed his own views as to the refractive index, specific gravity and viscosity, all of which should be as high as possible.As to heating tests, he agreed with Mr. Chapman’s criticism of the American method of heating to a very high temperature in a shallow dish, but he had found it useful to make a test by heating about 5 ounces of the oil in a dish of standard size at a temperature about looo F. lower than in the American test, and noting the time that it took to reach the condition in which the oil was used in the factory. The PRESIDENT asked whether Chinese and Japanese wood oils were ever mixed together. If this were done it would probably render the detection of adulterants more difficult. Mr. CHAPMAN, in reply, said that the difficulty of detecting adulterants would certainly be increased if the two kinds of oil were mixed, but he did not think this was done to any extent. He quite agreed with Mr. Hehner’s remarks as to the value of the hexabromide test, and also with the suggestion as to the influence of small quantities of impurities on polymerisation. The oil extracted with petroleum ether polymerised much less readily than oil prepared by expression in the ordinary way. The Chinese oils, however, did vary appreciably in their polymerising power, so that heat tests based upon comparison with a single standard oil must be quite useless for indicating such small amounts of adulteration as were sometimes reported. The iodine tiest discovered by Mr. Jenkins was an excellent supplementary test. With regard to Mr. Bolton’s question, he should not like to say positively that no genuine sample yielded a lower value than those he had given, but he thought that the figures he had recorded might be regarded as representing the limits which in the present state of knowledge it would be safe to adopt. He had met with lower iodine values, but was not sufficiently certain of the origin of the samples to include them. D

ISSN:0003-2654    

DOI:10.1039/AN9123700543

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

554 LITTLE : THE ESTIMATION OF MANGANESE THE ESTIMATION OF MANGANESE BY THE BISMUTHATE METHOD. BY H. F. V. LITTLE, A.R.C.S., B.Sc. (Read at the Meeting, November 6 , 1912.) THE bismuthate method for the estimation of manganese was originated by Schneider (Ding2. poly. J., 1888, 269, 224), and improved by Reddrop and Ramage (J. Chern. Soc., 1895, 67, 268) ; later it was modified by Ibbotson and Brearley (Chem.News, 1900, 82, 269 ; 1901, 84, 247, 302 ; 1902, 85, 59), and by Blair (J. Arner. Chem. Soc., 1904, 26, 793). The method consists in oxidising the manganese-present as man- ganous salt-in a cold solution free from chloride, and containing 25 per cent. by volume of nitric acid of sp. gr. 1-42, to the state of permanganate by agitation with an excess of sodium bismuthate. The solution is filtered through asbestos and titrated.Reddrop and Ramage added a slight excess of hydrogen peroxide, and titrated back with permanganate; but since in the presence of any appreciable quantity of ferric salt the use of peroxide is attended by an error proportional to the excees of reagent used, Ibbotson and Brearley employ ferrous sulphate. Blair does the same, and only differs from the others in the strengths of the standard solutions he employs. According to the writer’s experience, there is no advantage gained by using reagents of other than decinormal concentration, and he prefers to carry out the method as recommended by Ibbotson and Brearley, and described in Arnold and Ibbotson’s ‘( Steel Works Analysis,” diluting the oxidised solution, however, before filtration with an equal volume of 3 per cent.nitric acid. The object of the present note is to record a number of observations on the degree of accuracy of the method, which has been adversely commented upon by Brinton (J. Ind. and Eng. Chem., 1911, 3, 237, 376). According to this author, if the values of the standard solutions employed are deduced in the usual manner and calculated to manganese, the results are appreciably low, and the only way to obtain correct results is to determine an empirical faotor by standardisation against a known amount of manganese.So far as the estimation of small amounts of manganese is concerned, Brinton’s statement has been refuted by Hillebrand and Blum (J. Ind. and Eng. Chem., 1911, 3, 374), and a number of experiments have convinced the writer that the method gives admirable results when large percentages of manganese are concerned and the theoretical factor is employed. Firstly, evidence is supplied by the results obtained by E.Cahen and the writer (ANALYST, 1911, 36, 52) in determining the manganese in a number of ferros and pyrolusites by four different methods, for the theoretical factor was employed in working out the bismuthate results.Secondly, the test analyses given in Table I. lead to the same conclusion. The results given in Tables I. and 11. were obtained as follows : An approxim- ately FD permanganate solution was prepared from twice recrystallised potassium permanganate and distilled water, measured portions were reduced in nitric acid solution by a slight excess of concentrated hydrogen peroxide, and the oxidation of the solution with bismuthate effected as usual.A measured excess of approximatelyBY THE BISMUTHATE METHOD 555 & ferrous sulphate or hydrogen peroxide was added, and the excess titrated with the original permanganate solution, against which the ferrous salt and peroxide were ohecked.In each experiment marked with an asterisk the oxidised solution was filtered into an empty flask, and the titration subsequently performed. Low results are thus obtained. In the other experiments the oxidised solution was filtered directly into the ferrous salt or peroxide. The results are excellent. TABLE I. Manganese used in each experiment = 25 C.C. KMnO, solution.25 C.C. FeSO, solution = 30.78 C.C. KMnO, solution, Error. FeSO,, KMnO,. Manganese found c.,c. used. C.C. used. (in C.C. Original Solution). 25 5.83 24-95 - 0.05 25 5.83 24.95 - 0.05 25 5.77 25-01 4- 0.01 5.88 24-90 - 0.10 "25 5-92 24.86 - 0.14 35 18.06 26.03 + 0.03 :q5 TABLE 11. Manganese used in each experiment = 25 C.C. KMnO, solution. 25 C.C. H,O, solution = 27.97 C.C. KMnO, solution.H232. C.C. used. 25 25 25 35 35 *25 *25 KMnO,. C.C. used. 2.95 3.00 3-00 14-17 14 *18 3-08 3.10 Manganese found (in C.C. Original Solution). 25.02 24-97 24.97 24.99 24.98 24.89 24-81 TABLE 111. KMnO, used in titrations = 0*10174Mn. Excess of FeSO, over KMnO,. 1 C.C. = 0*001119 grm. Mn. Manganese. Fiund. Iron absent. ... 0.01125 ,.. 0.01119 ... 0.01125 Iron present. ... 0.01135 ...0.01182 ... 0.00445 .,.. 0.00447 Takk. 0.01128 0*01128 0.01128 0901128 0.01128 0*00450 0.00450 Error. 4- 0.02 - 0.03 - 0.03 - 0.01 - 0.02 - 0.11 - 0.13 Error. -0.03 mgrm. -0-09 ,) -0.03 7 7 - 0-03 mgrm. -0.06 9 , -0.05 7 7 -0.03 9 ,556 ESTIMATION OF MANGANESE BY THE BISMUTHATE METHOD In the experiments on smaller amounts of manganese, recorded in Table III., the use of hydrogen peroxide was discontinued.Known quantities of manganese were obtained by reducing a weighed amount of pure potassium permmganate, diluting to a known volume, and taking measured amounts of this solution. The approximately & permanganate used in the titrations was standardised against pure sodium oxalate, and the ferrous sulphate against the permanganate. In those experiments indicated as having been performed with iron present, sufficient ferric nitrate was added to contain 1 grm.of iron, and in each case under “manganese found ” a correction has been introduced (0*00031) for the manganese present in the ferric nitrate and determined by a blank experiment. The results are a trifle low-on an average 0*00005 grm.-which corresponds to only 1 drop of & permanganate.” With small quantities of manganese, it is a, matter of indifference whether the ferrous sulphate is added after filtration or the solution filtered into an empty flask.The error in these latter experiments only represents 0.005 per cent. Mn, taking 1 grm. of alloy, and duplicate experiments in ordinary working do not agree any closer than this. For example, three estimations carried out at the same time on the same manganese steel gave the writer 0-764, 0.756, 0.757 per cent.Mn respectively, and other alloys gave 0.259, 0.260, 0.271 per cent. Mn, 0.245, 0.236, 0-243 per cent. Mn, and 0.210, 0.203, 0.206 per cent. Mn. It should be mentioned that the writer has on several occasions encountered the curious experience which led Ramage (Chern.News, 1901, 84, 269) to reject the use of ferrous sulphate, and is at a loss to account for it. Neither Ibbotson and Brearley nor Blair refer to this matter. In the opinion of the writer, the bismuthate method is the neatest and most accurate of the volumetric methods that have been proposed for the estimation of manganese. ROYAL COLLEGE OF SCIEX’CE, SOUTH KENSINGTON, S.W.The President writes as follows: The bismuthate method of estimating manganese has been in use in my laboratory for seventeen years, chiefly in connection with the analysis of steel and iron. It is certainly the neatest method I am acquainted with, one of the most rapid, and, when carefully carried out, i s very accurate. We use Ibbotson and Brearley’s modification. The following results, in which the same steels were analysed by the bismuthate process and by Ford’s volumetric method, show how closely the two methods agree : Manganese Carbon.Ford’s Method. Bismuthkte Method. Steel casting ... ... 0.28 1 -508 1.500 1.503 1.183 Tyre ... .. . ... 0.41 1.185 1.191 * This is probably due to the presence of a little water (about 0’4 er cent.) in the recrystallised permanganate employed (Bl~un, J. Amer. Chem. s’oc., 1912, 34, 1379). Byurn’s study of the bismuthate method appeared some time after this paper WRS written.FOOD AND DRUGS ANALYSIS Manganese. 557 I FOG'S ~ Bismuthhate Carbon* Method. Method. Forging .., *.. ... 0.27 1.165 1.171 1.170 Rail ... ... ... 0.38 0.993 0.992 0.996 0.993 Crank axle ... ... 0.20 0.485 0.490 0.484 0.488" 0.480" 1 9 9 9 ... ... 0.27 0,834 0.844 0.841 The results marked with an asterisk (*I were obtained by Mr. J. H. B. Jenkins, who ttnalysed the samples in duplicate with me.

ISSN:0003-2654    

DOI:10.1039/AN9123700554

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

FOOD AND DRUGS ANALYSIS 557 ABSTRACTS OF PAPERS PUBLISHED IN OTHER JOURNALS. FOOD AND DRUGS ANALYSIS. Detection of Adulteration by Colloidal Chemical Methods. E. Marriage. (Zeitsch. Chenz. Ind. Kolloide, 1912, 11, 1-5; through Chem. Zentralbl., 1912, II., 1401.)-For the detection of cheap fruit-jelly, such as apple and gooseberry, and also of agar-agar in more expensive jams, a little of the sample is mixed with potassium iodide, and after cooling is covered with lead nitrate solution.From the appearance of the layers of lead iodide in the test-tube it is possible to draw conclu- sions as to the purity of the jam. Or a drop of the jam, to which 5 per cent. of potassium iodide has been added, is placed in a cell of 18 mm. diameter on a warmed object glass, and covered with a cover-glass of 16 mm.diameter: After cooling, the cell is filled up with 20 per cent. lead nitrate solution, and sealed with a large cover- glass, and the forms assumed by its contents examined under the microscope after the lapse of twelve to twenty-four hours. C. A. M. Detection of Aniline Dyes and Turmerie in Mustard. E. Sievers. (Zeitsch. Untersuch. Nahr. Genwsm., 1912, 24, 393-394.)-Although the method described by Bohrisch for the detection of turmeric (ANALYST, 1904, 29, 372) is reliable, the twelve hours’ extraction with aloohol renders it tedious, and the author, therefore, recommends the following method of procedure: A few grms.of the mustard are moistened with a drop or two of alcohol in a test-tube, and about 10 C.C. of ether are then added.The mixture is shaken, allowed to settle, and the ethereal layer is poured on a strip of filter-paper, and tested in the usual way for boric acid with turmeric. Bohrisch’s process for the detection of aniline dyes (Zoc. cit.) is simple and trustworthy. w. P. s.558 ABSTRACTS OF CHEMICAL PAPERS Detection of Minute Traces of Boric Acid by Means of Tincture of Mimosa Flowers.L. Robin. (Eighth Int. Coy. App. Chem., 1912, vol. I., 429- 432.)-Tincture of mimosa flowers is a more sensitive reagent than turmeric for the detection of boric acid. The reagent is prepared by digesting the flowers with 20 parts of 95 per cent. alcohol for ten minutes at the temperature of the boiling water bath, allowing to cool, and filtering. Since organic acids interfere, it is first necessary to incinerate such materials as milk, wines, etc.The liquid to be tested is evapor- ated in a 2-inch porcelain, flat-bottomed dish, and the residue is burnt and moistened with a few drops of water. A few (2 to 4) drops of the reagent are added, and then 5 per cent. hydrochloric acid drop by drop until the liquid loses its yellow colour. On evaporation to dryness, the residue is citron yellow if boric acid is absent, and its colour is unchanged on treatment with ammonia. In presence of boric acid the residue has a yellowish-brown colour, which changes to red on moistening with dilute (1 : 3) ammonia.A distinct pink is shown by less than 0*0005 mgrm. of boric mid. This test shows that boric acid is not merely an occasional, but a normal constituent of wines, and that it is present in minute traces in many samples of genuine milk.G. C. J. Detection and Estimation of Formic Acid in Syrups. H. Kreis. (Mitt. Lebensmittelimters. u. Hyg., 1912, 3, 205-210 ; through Chem. Zentralbl., 1912, II., 1312-1313.)-The author criticises the official Swiss method, according to which 100 grms. of the substance are acidified with sulphuric acid and distilled until at least 100 C.C.of distillate are collected. According to him, it is not possible to ascertain qualitatively whether, in the case of an acid saccharine liquid, formic acid has been added or not ; this can only be determined by a quantitative estimation of the formic acid, bearing in mind that formic acid may be produced in the course of the operation.The quantity of this latter depends on the manner of heating, the concentration of the liquid under distillation, and its acidity. The prescribed addition of dilute sulphuric acid should be omitted. If from 100 grms. of syrup, without the addition of acid, about 1 litre of distillate be driven over with a current of steam in the water-bath, it may safely be concluded that a quantity of formic acid exceeding 5 mgrms.cannot be due to the decomposition of the sugar, and that formic acid was originally present. The distillate obtained should be neutralised with calcium carbonate, being boiled under a reflux condenser with 1 to 2 grms. of finest precipitated chalk for ten minutes before it is evaporated down. J. F, B, Fractional Saponifleation of Fats.A. J. J. Vandevelde. (Ann. FaZsiJ, 1912, 5, 417-421.) - Results of an investigation regarding the effect of partial saponification of fats are recorded, statements having been made that such treat- ment alters considerably the character of certain fats which are employed as adulterants of butter. Coconut oil, lard, and butter fat were saponified wibh one-half the quantities of sodium hydroxide required for their complete saponifica- tion ; the unsaponified fats were then separated, an examination being also made of the insoluble non-volatile acids obtained from the soaps.The chief differencesFOOD AND DRUGS ANALYSIS 559 were noted in the mean molecular weights of the insoluble non-volatile fatty acids, particularly in the case of lard and butter fat. The molecular weights found were as follows: Fatty Acids from Fatty Acids from the Fatty Acids from the Original Fat.Non-saponified Portion. Saponified Portion. Coconut oil ... 208 211 214 Lard ... ... 274 273 240 Butter fat . , . 259 260 235 In the case of coconut oil, the saponification value was diminished from 271 in the original oil to 254 in the unsaponified portion, and from 236 to 219 in the case of butter fat, whilst the values for lard remained practically constant.The refractive indices and critical temperatures of solution were also lowered. w. P. s. Experiments on the Value of Ordinary and Treated Sawdust for Animal Nutrition. F. Honcamp, B. Gschwender, H. Mullner, and M. Reich. (Landzoirths. Vers.-Stat., 1912, 78, 87-114.) - An account is given of previous experiments by various authors, all of which tend to show that the fodder value of brushwood, sawdust, and wood-meal, whether raw or treated in various ways- e.g., by fermentation, etc.-is extremely small.It is true that Lehmenn had found that the fine subdivision of the wood, as in the case of wood-meal, slightly increased the digestibility of its components, but for practical purposes this effect was negligible.Pfeiffer, moreover, had shown that sawdust, besides affording no nutrition itself, also lowered the degree of utilisation of other fodder stuffs administered simultaneously. The subject has recently been reopened by the use of a product derived from sawdust by hydrolysis with sulphurous acid under pressure, with the formation of 20 to 25 per cent.of sugar. This product, obtained mainly from coniferous wood, occurs in the form of a brown meal, which is readily reduced to a fine dust by rubbing. It is utilised in admixture with molasses, for which it forms a solid vehicle, possessing certain physical advantages. The authors’ experiments were carried out upon sheep with the raw sawdust and with the treated product derived from it.These were administered in admixture with a standard ration composed of hay, cottonseed cake, and brewers’ grains molasses. Proximate analyses of the two wood materials, calculated on the dry substance, gave the following results : Non-nitrogenous Fat. Crude Fibre. Ash. Protein. Extract. Per Cent. Per Cent. Per Cent. Per Cent. Per Cent. Raw sawdust ...0.88 28-63 1-25 68.64 0.60 Treated sawdust 0.67 40.30 1.00 57-33 0.70 So far as the raw sawdust was concerned, the results of the present trials confirmed the observations of Pfeiffer. Not only was it without appreciable nutrient value in itself, but in certain directions it unfavourably affected the assimilation of the other components of the ration. The results obtained with the treated sawdust were slightly better in so far as the non-nitrogenous extract and the fat were concerned.The meal showed, on the dry substance, 26.20 and 0.73 per cent. respectively of these constituents assimilated. The quantity of fat introduced by the wood-meal is560 ABSTRACTS OF CHEMICAL PAPERS so small that its value is negligible. In the case of the non-nitrogenous extract, the improvement may be attributed to the saccharification of the wood fibre by the chemical treatment.Against this positive nutritive value must be set the fact that not only are the protein and fibre of the modified wood-meal totally unassimilable themselves, but the inclusion of this meal in the diet decreases the assimilation of the protein and fibre of the other constituents of the fodder.A further series of trials was made with a commercial fodder composed of the treated wood-meal end molasses. After correcting the results for the value of the molasses contained in the preparation, the conclusions arrived at in the case of the wood-meal itself were again confirmed. Thus the treated sawdust cannot be regarded as a satis- factory adjunct to molasses fodder, and more suitable vehicles must be sought for.J. F. B. Determination of Water in Molasses. C. C. Roberts. (Eighth Int. Cong. App. Chem., 1912, vol. 8, 53-37.)-Various methods of drying molasses for the estimation of moisture have been tested by the author. The direct drying of undiluted molasses in the air-oven at 98' to 99" C. gives low results owing to occlusion of moisture, and after a time the sample begins to gain in weight.The same remark applies when the sample is dried in a current of hydrogen. Drying the undiluted sample in air at 98" C. can only be performed when particular care is taken to place the sample drop by drop on the dish. Drying the diluted molasses on pumice is tedious, and yields low results. The author prefers to dry the diluted sample on a filter-paper coil, such as is used for the estimation of fat in milk by Adams' process.The water can be determined practically as exactly as in hydrogen and in a shorter time than by other methods. The sample is diluted to a uniform strength corresponding to about 30 per cent. of molasses. About 7 grms. of the solution, weighed by difference, are distributed as evenly as possible upon a dried coil of filter-paper, specially prepared for Adams' method, of dimensions 65 x 565 mm.The coil is dried for three hours in a water-oven, cooled and weighed. The loss is calculated as total water, including water of hydration of the sugars. If only the actual moisture is required, the sample must be dried at 70" C. to constant weight. J. F. B. Polarimetrie Estimation of Banana Starch. G. Baumert. (Zeitsch. Untersuch. Naltr. GenzLssm., 1912, 24, 449-452.)-The methods described by Lintner (ANALYST, 1897, 32, 179) and Ewers (ANALYST, 1908,33, 101, 481) may be employed for the estimation of this starch. Under the conditions of the former method banana starch has a specific rotation of +209*8', whilst in Ewers' method this value is +196*4'. The finely divided banana substance should be extracted successively with cold water, alcohol, and ether before the starch is estimated. (See also ANALYST, 1911, 36, 588.) w. P. s.

ISSN:0003-2654    

DOI:10.1039/AN9123700557

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

BACTERIOLOGICAL, PHYSIOLOGICAL, ETC. 561 BACTERIOLOGICAL, PHYSIOLOGICAL, ETC. Substitute for the Petri Dish for Cultivating Fungi and for Biological Analysis. P. Lindner. ( Wochenschr. f. Brau., 1912, 29, 589-590.)-The ordinary Petri dish presents certain difficulties in manipulation, particularly as regards infection from the air at the time of inoculation or when opening the dish to take off sub-cultures or specimens for examination.The vessels known as Lindner’s fungus- culture glasses are far more suitable for general purposes than Petri dishes. These glasses resemble in shape the inverted specimen jars, consisting of a round-ended cylinder with constricted neck, terminating in an enlarged foot. They are constructed of thin glass and are without stoppers, the opening in the foot being closed by a plug of cotton-wool.They are practically incapable of becoming infected from the air if properly manipulated, and the author has taken as many as sixty sub-sowings from one culture without infection. The walls are coated with an extremely thin layer of solid culture medium by rolling them round whilst the medium is fluid. The layer ~f medium and the glass both being extremely thin, the colonies are easily in- spected.Contrary to the Petri dish, it is the underside of the colony which is seen in these glasses, and the peculiar ring structure of the growth is visible unim- peded by surface vegetation. Further, by binding a sheet of gaslight photographic paper around the cylinder, and exposing it to the light, a direct photographic record of the culture at any stage is easily obtained.J. F. B. Estimation of Fat and Lipoids in Blood, and Notes on the So-called Lipolysis. L. Berczeller. (Biochem. Zeitsch., 1912, 44, 193-200.)-Results of experiments carried out by the author confirm the statement made by Shimidzu (cf. ANALYST, 1910, 35, 527), that more fat is obtained by extracting the blood with alcohol, and saponifying this extract, than is found by saponifying the blood directly according to the Kumagawa-Suto method.I t is recommended that 50 C.C. of blood be added drop by drop to 800 C.C. of alcohol ; the mixture is then boiled for five hours, filtered while hot, and the precipitate washed with alcohol and ether. The filtrate and wmhings are evaporated to about 100 c.c., and then saponified as described by Kumagawa and Suto (ANALYST, 1908, 33, 362).Horse’s blood yielded from 0.3 to 0.4 per cent., and dog’s blood from 0.5 to 0-6 per cent., of fat by this method, results which were about 0.1 per cent. higher than those obtained by the direct saponifica- tion method. The loss of fat (lipolysis) which has been observed when blood is aerated is probably due to the formation of compounds of the fat with other con- stituents of the blood; these compounds are soluble in alcohol, but not in ether.Alterations in the structure of the lipoids may also take place. I t does not seem probable that the loss is due to the action of bacteria. w. P. s. Polarimetric Estimation of Glucosamine in Ovomucoid and Pseudomucin, C. Neuberg and 0.Schewket. (Biochem. Zeitsch., 1912, 44, 491-494.)-A method which has been described by Neuberg and Ishida (ANALYST, 1912, 145j for the estimation of sugars in the presence of amino acids, peptones, etc., may also be employed for the estimation of glucosamine. I t depends on the fact that amino acids,562 AESTRACTS OF CHEMICAL PAPERS peptones, and albumoses, are precipitated by mercuric acetate and phosphotungstic acid, whilst sugars and glucosamine in neutral solution are not precipitated by these reagents. One grm.of ovomucoid is hydrolysed by boiling for three hours with 50 C.C. of 7 per cent. hydrochloric acid, the solution is then evaporated, and the residue is dissolved in 13 C.C. of water. A 25 per cent. mercuric acetate solution is now added in quantity sufficient to make the total volume 25 ccc.After filtration 12.5 C.C. of the filtrate are treated with 2.5 C.C. of a concentrated phosphotungstic acid solution, the mixture is again filtered, and the filtrate examined in the polariscope. In the case of pseudomucin the hydrolysis with hydrochloric acid should be continued for thirty hours. Ovomucoid was found to contain 24 per cent., and pseudomucin 36 per cent., of glucosamine.w. P. s. Detection of Combined Glycuronic Acid in Normal Urine. (1. Neuberg and 0. Schewket. (Biochem. Zeitsch., 1912, 44, 502-504.)-The naphthoresorcinol test described by Tollens (ANALYST, 1908, 33, 326) is a general reaction for carboxylic acids, and the orcinol test for glycuronic acid also gives a reaction with pentosans.Both these reactions, however, may be used for the detection of combined glycuronic acid if they are applied to the ethereal extract of a urine, and not to the latter directly. The test is carried out as follows : Ten C.C. of the urine are acidified with 2 O.C. of dilute sulphuric acid, 10 C.C. of alcohol are added, and the mixture is shaken out with 20 C.C. of ether. The ethereal extract is washed with water, filtered, and evaporated after the addition of a small quantity of water.The residual aqueous solution is then divided into two portions, to which the tests are applied. The urine to be tested should be as fresh as possible ; when urine is kept, the combined glycuronic acid is decomposed, with the formation of free glycuronic acid which is not soluble in ether.w. P. s. The Indole Reaction. H. Zip€el. (Zentralbl. Bakt. Parasit., 1912, 64, 65-80 ;. through Chem. Zelztralbl., 1912, II., 1497.)-The possibility of the production of indole by bacteria depends on the presence of the tryptophane group (indole-a-amino- propionic acid) in the nutrient medium. The author recommends the use of a, medium prepared by dissolving 5 grms.of asparagin, 5 grms. of ammonium lactate, 2 grms. of dipotassium phosphate, 9.2 grms. of magnesium sulphate, and from 0.1 to 0.5 grm. of tryptophane, in 1 litre of water. This solution has the advantage over peptone solution in being free from colour ; it is also con&ltnt in composition. Any of the ordinary tests may be employed for detecting the formation of indole by bacteria in a tryptophane medium, but Bohme’sp-dirnethylaminobenzaldehyde reaction.gives the best quantitative results. w. P. s. Occurrence of Mercury in the Hair of Persons who have received Subcutaneous Doses of Mercury Compounds. Detection of Minute Quantities of Mercury. C. Strzyzowski. (Chem. Zed., 1912, 36,1237-l239.)-By means of the following method the author has detected traces of mercury in the hair of persons who had previously been subjected to treatment with mercury compounds. The method is capable of detecting the presence of 1 part of meroury in about 90,000,000BACTERIOLOGICAL, PHYSIOLOGICAL, ETC.563 parts of substance. From 2 to 10 grms. of the h d r are washed successively with ether, alcohol, and hot water, and are then digested with hydrochloric acid, small quantities of potassium permanganate being added from time to time until the substance has dissolved.The solution is filtered, evaporated partially, and then treated with hydrogen sulphide ; after the lapse of twenty-four hours, the precipitate, consisting of sulphur and traces of mercury sulphide, is collected on a filter, then dissolved in a small quantity of hydrochloric acid and potassium chlorate, the solution is filtered, and the filtrate is evaporated to about 1.5 C.C.This residual solution is diluted, filtered, and a few small strips of copper-foil are added to the filtrate, which is then boiled for a few minutes; the copper-foil is now removed, dried, and transferred to a tube, the upper end of which is then drawn out to a capillary.After the air has been exhausted from the tube (while the lower part of the tube is heated to about 130° C.) the capillary is sealed, and the tube is heated over a flame in order to volatilise the mercury from the foil into the capillary. The latter is then examined under the microscope, when minute globules of mercury will be observed. These globules may be rinsed with alcohol from the capillary, and treated on a microscope slide with a trace of iodine.A characteristic action of the iodine on the globules may be observed under the microscope. w. P. s. Toxicity of Mushrooms : Their Haemolytic Aetion. J. Parisot and Vernier. (Compt. rend., 1912,155, 620-623.)-The symptoms observed in cases of poisoning by mushrooms point to haemolysis BY the cause of death.This has been confirmed, both in vitro and in vivo, by experiments made with extracts of fungi in sodium chloride solutions containing 9 grms. of salt per litre. Examination of the blood of animals killed by the injection of these extracts showed a cherry-red coloration of the serum (hamoglobinamia), and the globules were haemolysed in saline solutions. The Amanita phalloides, Fr., was found to possess the most powerful hemolytic action, but all poisonous species had an intense, though less marked, action. Moreover, many of the edible species also exert a destructive action on the blood, notably Amanita rubescens, Pers., Hydnum repandurn, Lin., Tricholomia nudum, Fr., Laccaria Iaccata, Be.and Br., Craterellus cornucopioides, Fr., etc. Certain species which do not cause haemolysis when young and fresh will develop this property as they get older.Generally the hamolytic action is attenuated by heating, but in certain cases a prolonged period of heating at a high temperature is not sufficient to destroy it entirely. The authors’ experiments show that the hamolytic action is due to a glucoside. The activity of the glucoside is very considerably weakened and retarded by certain substances, such as milk, and par- ticularly cholesterol (yolk of egg).When the haemolytic action of the fungus extracts had been destroyed by heating, it was found that their toxic action was either very much attenuated or eliminated. When subjected to the ordinary operations of cooking, the mushrooms frequently retain their hamolytic activity to a certain extent, and it is recommended that such fungi should be cooked for a prolonged time at a high temperature, preferably with the use of oil, in order to effect the complete hydrolysis of the haemolytic glucoside. J.F. B.564 ABSTRACTS OF CHEMICAL PAPERS Estimation of Oxalic Acid in Vegetable Substances. A. Gregoire and E. Carpiaux. (BUZZ.SOC. Chim. BeZg., 1912,26, 431-434.)-Oxalic acid, in the form of its calcium, potassium, and sodium salts, is found widely distributed in the vegetable kingdom, and to a much less extent in the animal kingdom. At the same time this acid is but rarely estimated in the ordinary course of analysis of vegetable foodstuffs, and the authors therefore describe a method for its estimation, this method having been applied particularly to such substances as sesame cakes and other products, which also contain large quantities of phosphates, calcium salts, and proteins. Five grms.of the finely powdered, fat-free cake are heated on a water- bath for one hour with 20 C.C. of 4 per cent. hydrochloric acid ; a quantity of sodium sulphate sufficient to precipitate the calcium salts present is then added, and after cooling the mixture is mixed with about 100 C.C.of 94 per cent. alcohol. The precipitate is collected on a filter, washed with alcohol, a slight excess of ammonia is added to the filtrate, the alcohol is removed by evaporation, s and the solution is acidified with hydrochloric acid and filtered. Calcium acetate is added to the filtrate, the calcium oxalate is collected on a filter, washed, dissolved in hydrochloric acid, and the solution evaporated nearly to dryness.This residue is acidified with a few drops of 25 per cent. sulphuric acid, and sufficient anhydrous sodium sulphate is added to form a dry mass, which is then extractod six successive times with ether. The ethereal extracts are rendered ammoniacal, evaporated, the residue is dissolved in water, and the oxalic acid is precipitated with calcium acetate, collected, ignited, and weighed.A sample of sesame cake examined yielded a qusntity of oxalic acid, corresponding with 2.94 per cent. of crystallised calcium oxalate (CaC,O,.H,O). w. P. s. Estimation of the Activity of Commercial Rennet. W. van Dam. (Landw. Vers.-Stat., 1912, 78, 133-163 ; through Ghern.Zeatralbl., 1912, II., 1313- 1314.)-The author has observed that in presence of a low hydrogen ion concentra- tion (2.37 x 10- 5N> the digestive activity of rennet towards paracasein runs exactly parallel to its coagulating activity. He has based on this an accurate method for establishing the activity of a normal rennet, and he proposes as a standard 1 : 100,000 rennet powder, a preparation which yields digestion products, under the conditions prescribed, which are equivalent to 26% C.C.of & acid in Kjeldahl's nitrogen estimation. The digestion experiment is carried out as follows : Into a 100 C.C. flask is weighed as much purified casein as corresponds to 0.8 grm. of anhydrous substance (Kjeldahl factor 6.37). To this are added 20 C.C.N-acetic acid, 10 C.C. N-sodium hydroxide, and 70 C.C. of boiled water, whereby the above-mentioned concentration of hydrogen ions is established. The flask is slowly rotated in the thermostat at 30" C. After settling, 1 C.C. of a 1 per cent. solution of the rennet powder is added. A control test is also set up containing 1 C.C. of the rennet extract previously heated at 90" to 100" C.for ten minutes. The digestion is continued for twenty-four hours at 30' C. ; 75 C.C. are then filtered off for the Kjeldahl estimation, and the number of C.C. of acid neutralised represents the digestion value. By means of the square root law, according to which the quantity of casein digested is proportional to the square root of the enzyme concentration, it is calculated whatBACTERIOLOGICAL, PHYSIOLOGICAL, ETC.565 concentration of the rennet powder must be taken to give a result approximating to 26.5 C.C. of acid. This solution must then be compared with one or more calf- stomach infusions, in order to verify the parallelism of the digestive and coagulating functions. The neutralised infusions of calf - stomach are diluted until their coagulating power is approximately equal to that of the rennet powder solution, and the digestive activity of these infusions is then estimated.If the digestion and coagulation are found to be sufficiently parallel, the solution of rennet powder is again tested for its digestive activity, and, provided the result does not differ by more than 2 C.C. from 26.5 c.c., the strength of the powder may be calculated by the square root law on the basis of 26.5 C.C. for a standard powder of 1 : 100,000 in 1 per cent.solution for 75 C.C. of digested products. J. F. B. Colorimetric Method for the Estimation of Tryptophane, and the Quantity of Tryptophane in Horny Tissue and Other Proteins. H. Fasal. (Biochenz. Zeitsch., 1912, 44, 392-401.)-The method is based on the violet coloration which is produced when tryptophane, or a substance containing it, is treated with concentrated sulphuric acid and glyoxylic acid. The coloration obtained is compared with that produced by known quantities of pure tryptophane. The author finda that the use of glyoxylic acid in the test is preferable to that of other aldehydes. In testing horn substance, horny epidermis, etc., the material was previously subjected to peptic and pancreatic digestion, and extraction with alcohol and ether. The following quantities of tryptophane were found in various materials : Horny epidermis, 0.30 per cent.; horn (ox), 0.17 per cent.; human nails, trace; hair (white, blonde, and dark), none ; casein, 0.65 per cent. ; edestin, 0.38 per cent. ; lact- albumin, 3.07 per cent. w. P. s.

ISSN:0003-2654    

DOI:10.1039/AN9123700561

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

BACTERIOLOGICAL, PHYSIOLOGICAL, ETC. 565 ORGANIC ANALYSIS. New Method for Estimation of Alchohol. G. B. Frankforter and F. C. Frary. (Eighth Int. Cortg. App. Chern., 1912, vol. 22, 87-127.)-When sufficient potassium fluoride is added to aqueous alcohol, the liquid separates into two layers. If now water is added gradually, with shaking between each addition, a point is reached when the solution again appears homogenous.The end point can be determined very closely if a little solid spirit blue be present, since this makes a very thin film of alcohol easily visible. The analytical method consists in trans- ferring some pure, anhydrous potassium fluoride to a, weighed flask, weighing, adding some of the aqueous alcohol to be tested, reweighing, cooling to about 25O C., and adding water a drop at a time until the solution becomes homogenous, A final weighing of the flask is then made, and the alcoholic strength of its contents, and thus of the sample, calculated by reference to Table I.Reference to the table is necessary before commencing an analysis. For example, if a spirit believed to contain from 40 to 50 per cent. of alcohol is to be anslysed, the table shows that as much as 12 grms.of potassium fluoride may be necessary to insure separation into two layers if 100 grms. of spirit is taken for the test. If the quantities taken are exactly 12 grms. and 100 grms., and if 20 grms. of water is necessary to render the566 ABSTRACTS OF CHEMICAL PAPERS solution homogenous, then the final solution contains 12 grms.of potassium fluoride in 120 grms. solvent, or 10 grms. per 100 grms. solvent. By reference to the table, it is seen that this corresponds to 44 per cent. alcohol, This is the strength of the alcohol at the end of the experiment after the addition of 20 per cent. of water, and therefore the strength of the original spirit was 52.8 per cent. Potassium fluoride does not salt out methyl alcohol, but the above method gives accurate results with methylated spirit between 15 and 30 per cent. total alcoholic strength.With stronger or weaker methylated spirit the method gives results which may be in error by as much as 1 per cent. TABLE I. Grms. per 100 Grms. Solvent. KF. C2H,0H. 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 2.82 2.92 3.02 3-13 3.24 3.36 3.48 3.61 3.75 3.90 4.06 4.23 4.41 4.60 4.80 5.01 5.24 5-48 5-74 6.02 KF.C,H,OH. 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 6.32 6.65 7.01 7-40 7-83 8 -30 8-81 9.37 9.98 10.64 11-36 12.14 12-99 13.92 14.94 16.07 17.33 18.75 20.26 21-80 KF. C,H,OH. 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 23.4 25.1 26.9 28.8 30 *8 32.9 35.0 37-2 39.4 41 -7 44.0 46.4 49.0 51.9 55.1 58.7 62-8 67.5 78.6 Potassium fluoride absorbs 62 per cent.of its weight of water against 32 per cent. taken up by lime, and, probably on account of its greater solubility, its dehydrating action is more rapid than that of most drying agents. I t cannot, however, be used for the production of alcohol of a strength exceeding 97.5 per cent. Though potassium fluoride is the most suitable substance for use in the analytical method above described, potassium carbonate may be used.Very small amounts of insoluble matter render the end point uncertain, and it is therefore better to use a filtered, concentrated solution of known strength. The best indicator is solid phenolphthalein, as potassium carbonate bleaches spirit blue. Table 11. gives a guide as to suitable quantities for analysis and the means of calculating the results :ORGANIC ANALYSIS 567 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 TABLE 11.Grms. per 100 e m s . Solvent. C2H,0H. 5.0 5.2 5.4 5.7 6.0 6.3 6.6 7.0 7.4 7.8 8.3 8.8 9.3 9.9 10.5 11.2 11.8 12.5 13.2 14.0 14.9 15 -8 16.8 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2.5 2 C,H,OH.17.8 18.9 20.0 21.1 22-3 23.5 24.7 26-0 27.2 28.5 29.8 31.1 32.5 34-0 35.7 37.5 39.6 42.0 44.7 48.3 50-3 52.6 G. C. J. Detection of Methyl Alcohol in Ethyl Alcohol and Spirits. A. Bono. (Chem. Zeit., 1912, 36, 1171.)-Twenty-five C.C. of spirits and 50 C.C. of water (or 10 C.C. of strong alcohol diluted to 75 c.c.) are distilled in a flask with a delivery tube passing to the bottom of a, second flask containing 50 C.C.of a cold saturated solution of potassium bichromate with 60 C.C. of sulphuric acid per litre. The vapours issuing from this second &sk pass through a condenser into a graduated cylinder. The first 25 C.C. of the distillate, in which formaldehyde can rarely be detected, are rejected, whilst the next 1Q C.C. are tested for formaldehyde &B follows : Ten drops of a 0.5 per cent.aqueous solution of phenylhydrazine hydrochloride, 1 drop of 0.5 per cent, sodium nitroprusside solution, and 10 drops of 10 per cent. sodium hydroxide solution, are added to 2 C.C. of this fraction, which in the presence of formaldehyde gives a blue coloration changing to green and then to yellowish-red. About 1 per cent.of methyl alcohol may thus be detected. C. A. M. Analysis of Oil of Citronella. V. Boulez. (Bull. Xoc. Chim., 1912,11, 915- 917.)-From 25 to 50 grms. of the oil me treated with 100 (or 200) grms. of a,568 ABSTRACTS OF CHEMICAL PAPERS concentrated solution of sodium bisulphite prepared by saturating the ordinary com- mercial solution of sp. gr. 1-345 with sodium sulphite. The mixture is shaken until the citronella1 has combined with the bisulphite, after which it is allowed to stand for two to three hours, then diluted with 100 (or 200) C.C. of water, and heated with occasional agitation for several hours on the water-bath beneath a, reflux condenser.There should then be a sharp separation between the dissolved aldehydic part and the uncombined oily part of the sample. The latter is collected in a tared flask and weighed, and the difference between the result and the original weight of oil gives the amount of citronellal. The geraniol is then estimated in the uncombined portion by acetylation in the usual way.Four samples of citronella oil thus analysed gave the following percentage results : Citronella1 ... ... ... ... ...... 18-00 19-00 58.00 63-00 Geraniol ... ... ... ... ... ... 44-28 45.59 36.16 35.56 Citronella1 and geraniol expressed as geraniol ... 62.07 - - 81.50 I. 11. 111.. IV. C. A. M. Determination of Aldehydes. B. G. Feinberg. (Eighth Iizt. Cong. App. Chem., 1912, vol. 1, 187-203.)-An investigation concerning the accuracy with which formaldehyde, benzaldehyde, salicylaldehyde, p-hydroxybenzaldehyde, anis- aldehyde, and vanillin, may be estimated by methods depending on the use of bisulphite, neutral sulphite, or substituted hydrazines.The most accurate method for the estimation of benzaldehyde depends on the use of p-nitrophenylhydrazine. A convenient concentration of benzaldehyde is 0-25 grm. in 75 C.C. of 4 per cent. acetic acid. To this about twice the theoretical amount of the reagent is added, dissolved in 30 C.C.of 30 per cent, acetic acid. After five hours the precipitate is collected on a Gooch crucible, washed with 10 per cent. acetic acid, and dried ad 105' to llOo C. Each gram of precipitate should correspond to 0.440 grm, benzaldehyde. Under the conditions described the benzaldehyde will be under- estimated by 1.2 * 0-7 per cent.A similar method gives good results (error t 0.5 per cent.) with anisaldehyde and vanillin, but these aldehydes can be estimated with almost equal accuracy by means of p-bromphenylhydrazine, which is useless as a quantitative precipitant for benzaldehyde. The best results for salicyl aldehyde were obtained with p-nitrophenylhydrazine, but the error, though fairly constant, was large, namely-5 k 0.5 per cent.G. C. J. Titration of Azo Dyestuffs by Means of Hydrosulphite. E. Grand- mougin and E. Havas. (Chem. Zeit., 1912, 36, 1167-1169.)-A solution of an alkali hydrosulphite reacts with azo dyestuffs in accordance with the equation : R . N : N . R' + 2Na2S,04= RNH, + R"H,+ 4NaHS0,. The reagent is prepared by dissolving about 3 grms. of solid sodium hydrosuIphite in a litre of recently boiled and cooled water, to which 5 C.C.of 30 per cent. sodium hydroxide solution has been added, and allowing the solution to stand for one toORGANIC ANALYSIS 569 two hours. For the titration of the dyestuff the reagent is introduced into the flask, A , and the tube, B, I t should be used within four or five hours of its preparation. of the apparatus shown in the figure, whilst 50 C.C.of a 0.1 per cent. solution of the dyestuff is placed in the lower flask, where it is diluted to 100 or 150 c.c., and treated with 5 C.C. of dilute hydrochloric acid (1 vol. of 36 per cent. acid to 10 vols. of water). The apparatus is then closed, and the air expelled from it by a current of gas, which is subsequently lit at a burner connected with D, after which the burette is filled and the dyestuff titrated. The method can not only be used in comparative estimations, but will yield absolute values by standardising the reagent upon a sample of a dyestuff of known purity.The amounts of hydrosulphite used are inversely propor- tional to the molecular weights of the anhydrous dyestuffs, as is shown by the following examples : Molecular Weight.C.C. Orange 11. ... ... 28.0 350 Chrysoin ... ... 31.0 316 True Red (Echtritj ... 24.5 400 I n a series of estimations of five commercial samples of Orange 11. the standard was prepared by twice re- crystallising a sample of this dyestuff from a weak solution of sodium chloride, washing the crystals, and drying them at 130' to 140' C. until constant in weight.Taking this now to represent 100 per cent. of the pure dyestuff, the results obtained in com- parison with the commercial samples showed per- B 3 centages ranging from 75.2 to 89.3. SO far the method has only been applied to acid mono-azo dyestuffs, which on reduction are sharply transformed into colourless products. Modifications will be necessary to adapt the method to the examination of other azo dyestuffs, such as poly-azo dyestuffs, which only dissolve with difficulty, The accuracy of the method has been controlled by precipitation of the acid dyestuffs with basic dyestuffs and by colorimetric comparisons. C.A. M. Arsenic Xanthogenate and its Use in Analysis. N. Tarugi and F. Sorbini. (Boll. China. Farm., 1912, 51, 361-370 ; through Chem.Zentralbl., 1912 II., 1398-1399.) - Arsenic xanthogenate, As(SCSOC2H,),, is prepared by mixing a slightly alkaline solution of potassium arsenite with a solution of potassium xanthogenate, allowing the mixture to stand for fifteen minutes, acidifying it with acetic acid, washing the crystalline precipitate with water, and recrystallising it after drying over sulphuric acid. I t melts at 94.8' C.(corr.), is completely insoluble in hot and cold water, and does not volatilise with steam. It is sparingly soluble in alcohol,570 ABSTRACTS OF CHEMICAL PAPERS more soluble in ether, benzene, carbon bisulphide, and chloroform, and soluble in hot glycerol, vegetable oils, vaseline, and lanoline. When heated with aniline it is quantitatively decomposed into diphenylthiourea and arsenic trisulphide. Cold mineral acids do not attack it, whilst cold concentrated solutions of caustic alkali dissolve it unchanged.The reaction with aniline may be used in the estimation of minute quantities of arsenic, the separated arsenic xanthogenate being heated on the water-bath with a few drops of aniline, and the residue of arsenic trisulphide heated at 150" C.until constant in weight. Quantitative results are obtained, but in estimating minute traces of arsenic it is preferable to extract the arsenic xanthogenate with chloroform and to evaporate the extract. Conversely the method may be used in the analysis of potassium xanthogenate. For a volumetric estimation 10 C.C. of an alkaline solution of arsenic trioxide of known strength are vigorously shaken with 5 C.C.of the potassium xanthogenate solution and allowed to stand for fifteen minutes, after which the liquid is acidified with acetic acid, diluted with water, vigorously shaken with 20 C.C. of benzene to dissolve the arsenic xanthogenate, and allowed to settle. The aqueous layer is now made up to 100 c.c., 10 C.C. thereof neutralised with sodium bicarbonate, and the excess of arsenic titrated with TG iodine solution, each C.C.of which is equivalent to 0-024 grm. of potassium xanthogenate. Carbon bisuZphide: This may be detected by means of an arsenite solution containing aniline and alcohol. The solution under examination is added to this reagent, or the air is made to bubble through it for some minutes; the liquid then neutralised with dilute sulphuric acid and diluted with water, and the precipitate freed from arsenic eulphide by recrystallisation.Glistening white crystals of a-diphenylthiourea (melting-point, 151" C.) are obtained : CS, + 2C6H,NH, = H,S + (C,H,NH),CS. The reaction is not quantitative. A gravimetric method of estimating carbon bisulphide is to add the solution drop by drop to a solution of arsenic trioxide in potassium hydroxide solution containing alcohol, with continual cooling, or to pass the air to be tested through this reagent.After standing for fifteen minutes, the liquid is acidified with acetic acid, and allowed to stand for twelve hours, and the precipitate separated and weighed. Or carbon bisulphide may be estimated volumetrically with arsenite solution in an analogous manner to that used in estimating potassium xanthogenate.C. A. M. Composition and Properties of Glycin Picrate and Separation of Glycin from Alanin. P. A. Levene and D. D. van Slyke. (J. Biol. Chem., 1912, 12, 285 ; through Chem. Zentralbl., 1912, II., 1196.)-The formula of glycin picrate was found by analysis to be (C,H,O,N),. C6H30,N3 and not C,H,0,N.C6H,0,N3. It melts at 199-200" C., decomposes at 202' C., and is very soluble in hot water, while water at 0" C.dissolves only 1.7 grms. per 100 C.C. E'or the separation of glycin and alanin the mixture is dissolved in 3 to 4 parts hot water, picric acid added in quantity less than sufficient to neutralise both bases, but more than enough for the glycin alone, and the glycin picrate allowed to crystallise out at 0" C.0. E. M.ORGANIC ANALYSIS 571 Estimation of Glycerol in Fats. R. Willstatter and A. Madinaveitia. (Ber., 1912, 45, 2825-2828).-The method of Zeisel and Fanto, according to which the fat is saponified with alcoholic potash and the glycerol is afterwards converted into isopropyl iodide by the action of hydriodic acid, is generally considered to be the most accurate method for the estimation of glycerol.The authors now propose an improved method, whereby the fat is treated directly with hydriodic acid. About 0-2 grm. of the glyceride is weighed out into the decomposition flask of the Zeisel- Fanto apparatus and treated with 10 C.C. of hydriodic acid of sp. gr. 1.80. The mixture is heated to 100" to 115" C. (bath temperature) until reaction sets in, as shown by a strong separation of iodine and precipitation of the silver solution in the receiver.The temperature is maintained constant until the silver solution clears again (twenty to forty minutes) and the reaction is practically finished. Then the bath is raised to a temperature of 130" to 140" C., and heating is continued for at least an hour.This is not important in the case of triolein, but is necessary with tristearin. Results are satisfactory, e.g., 9.87 per cent. of glycerol for tristearin, against 10.33 calculated, and 10.32 per cent. for triolein, against 10.41 calculated. If hydriodic acid of sp. gr. 1.70 be employed the results are too low, especially when larger quantities of fat are taken for analysis.Note on the Acetyl Value of Fats.-As a measure of the hydroxy fatty acids and fat alcohols, Benedikt and Ulzer determine the acetyl value of the portion of saponified fat acid insoluble in water. Lewkowitsch (J. SOC. Chem. Ind., 1897, 16, 503) proposed instead to estimate the acetyl value of the fat direct without previous saponification. The authors have found that such treatment is not permissible, since acetic anhydride displaces the higher fatty acids in glycerides, with production of aoetins. J.F. B. The isopropyl iodide distilled over is weighed as silver iodide. Quantitative Estimation of Aetive Hydrogen in Organic Compounds by Magnesium-methyl Iodide. T. Zerewitinofl. (Ber., 1912, 45, 2384-2389.)- Hibbert (Chem. SOC. Trans., 1912, 101, 328) has stated that the lower alcohols of the fatty series, when treated with magnesium-methyl iodide in phenetole solution, yield considerably less than the theoretical quantity of methane.Working according to his standardised method for the estimation of hydroxyl groups or '' active hydrogen " (ANALYST, 1911,37,117) and using pyridine as the solvent, the author fails to confirm this statement and has obtained theoretical results from methyl, ethyl, propyl, and isobutyl alcohols.He has further extended the investigation of his method to the case of diamines, including representatives of the aliphatic, benzene, benzidene, stilbene, and naphthalene series. The estimations were carried out both in pyridine and anisole solutions, The results were concordant, and showed that substances containing two NH, groups react with two atoms of 6 6 active hydrogen" at the ordinary temperature, and with three atoms on heating at 85" C., but that the fourth atom does not react at all with magnesium-methyl iodide.Malonamide is an exception, reacting with four atoms on heating, but in the author's view one of these "active hydrogen " atoms is derived from the CHz group.The hydrocarbons of the indene and fluorene type in pyridine solution do not react at the ordinary temper-572 ABSTRACTS OF CHEMICAL PAPERS ature, but on heating at 8 5 O C. for five minutes, a quantity of methane is obtained corresponding to one atom of “ active hydrogen.” The same applies for the hydro- carbons derived from fluorene by the substitution of a phenyl or naphthyl radicalfor one hydrogen in the CH, group; on the other hand, the carbinols derived from these latter hydrocarbons behave as tautomeric bodies, reacting only partially at the ordinary temperature and with only one active hydrogen ” atom on heating.5. F. B. New Stability Test for Nitrocellulose Powders. S. A. Weirman. (Eighth Int. Cong. A p p Chem., 1912, vol.4, 157-159.)-The ability of a powder to take up and retain moisture is suggested as a measure of its stability, on the ground that this determines the amount of its hydrolytic decomposition. The sample, of which 10 grms. are sufficient, is weighed, dried for seven days in a sulphuric acid desiccator and weighed again; if is then exposed over water in a desiccator for forty-eight hours, and the gain in weight noted.The total range in weight divided by the surface area of the grain gives the actual “moisture range,” values of which for different sizes of grain are comparable. The results are found to be parallel with those of the potassium-iodide starch test. 0. E. M. Exudation of Nitroglycerine from Dynamite. C. P. Beistle. (Eighth Int. Cong, App. Chem., 1912, vol.4, 7.)-The dynamite is placed in a centrifugal machine, the cups of which describe a circle of 7-inch radius, and rotated for one minute at 600 revolutionsper minute. The results obtained agreed with those of the 40” C. oven tests whereas the compression method gives results dependent upon the nature of the absorbent used. 0. E. M. Separation of Nitroglycerine from Nitro-Compounds.A. L. Hyde. (Bighth Int. Cong. App. Chem., 1912, vol. 4, 69-76.)--Complete fraotional separation by two immiscible solvents, in one of which the nitroglycerine is more soluble than the nitro-compound, while in the other these conditions are reversed, could not be effected ; but an empirical method, in which acetic acid and carbon disulphide were the solvents chosen, gave a definite fraction of the nitro-compound, nearly free from nitroglycerine, in the carbon disulphide.The proportion recovered was dependent on the particular nitro-compound under investigation. 0. E. M. Effect of the Nitrotoluenes on the Determination of Nitroglycerine by Means of the Nitrometer. C. P. Storm. (Eighth Int. Cong. App. Chem., 1912, vol. 4, 117-125.)-The nitric acid liberated from nitroglycerine by sulphuric acid in the nitrometer is taken up quantitatively by admixed mononitrotoluene until this is all converted into dinitrotoluene, if there is sufiicient nitroglycerine.Hence, mixtures of these two substances containing less than 35.6 per cent. of nitroglycerine, yield no gas in the nitrometer ; with more than this amount the error is 0.55 grm.of nitro- glycerine per grm. of mononitrotoluene present. Pure dinitrotoluene and trinitro- toluene are without effect on the determination, but commercial liquid products cause an error proportionate to their content of mononitrotoluene. A liquid dinitro- toluene investigated caused an error of 0-0628 grm. nitroglycerine per grm., andORGANIC ANALYSIS 573 therefore contained 11.36 per cent.of mononitrotoluene. An explosive containing 25 per cent. of nitroglycerine and 10 per cent. of this liquid dinitrotoluene would thus show an error of 0-628 per cent. in its nitroglycerine content. 0. E. M. Modified Form of Stability Test. H. C. P. Weber. (Eighth Into Gong. Am. Chem., 1912, vol. 4, 147-156.)-The sample, if necessary cut into pieces weighing about 0.2 grm., is placed in a test-tube partly immersed in a glycerol or oil bath, and the time required to produce explosion measured with a stop-watch.The tempera- tures chosen for the bath in this investigation were 200°, B O O , and 160° C.; for each sample at least three tests were made at each temperature, and a curve plotted from the average values and the temperatures. The curves fell into three groups for the samples tested, corresponding to their known classification into good, fair, and bad.Stable powders gave a pronounced bend, and a ratio of explosion periods at 200" and 160" C. of at least 2 : 9, while the ratio fell to 2 : 3 with unstable powders, and the points did not fit a smooth curve so well. Nitrocellulose gave a flatter curve, more nearly corresponding to the theoretical curve constructed on the assumption that the reaction velocity doubles for every 10" C.0. E. M. Determination of Water in Petroleum and its Products. J. C. Allen and W. A. Jacobs. (Eighth Int. Cong. App. Chem., 1912, vol. 10, 17-23.)-A criticism of ten methods, all of which find use for the determination of water in petroleum or its products. The most exact is that depending on the volume of hydrogen evolved when the sample is treated with finely-cut sodium.The manipula- tion is identical with that of the well-known carbide method, the results of which are less accurate on account of the solubility of acetylene in oils. Of the other methods, the most accurate is direct distillation of a considerable quantity (20Q grms.) of the sample.This method is accurate to within 0.03 per cent, if the distillation is pushed to 150" C., the water removed from the receiver by a micro- pipette and weighed, and the condenser cleaned with a tared plug of absorbent cotton, which has been previously wetted and squeezed as dry as possible. Ice-water should be fed to the condenser to obtain results of the accuracy stated.The other methods described, though useful for some purposes of works control, are only approximately exact, and some of them are only appliable to heavy oils and greases. G. C. J. Conditions for the Complete Hydrolysis of Proteins. D. D. van Slyke. (J. Biol. Chzem., 1912, 12, 295 ; through Chem. Zentralbl., 1912, II., 1222.)-Various proteins were heated with 20 per cent.hydrochloric acid at 100" C., and also with 3 N. hydrochloric acid at 150' C., and the ammonia, total nitrogen, amino-nitrogen, and melanin determined in the reaction-product. The amino-nitrogen reaches a maximum, which is the same whether the heating is at 100' or 150" C., when the acid hydro- lysis of a protein is complete. The author was unable to confirm the statement of Henriques and Gjaldbaek (Zeit. physioZ.Chem., 1910, 67, 8 ) that egg albumin, when heated with 20 per cent. hydrochloric acid at 100' C., gives only 90 per oent. of the amino-nitrogen obtainable by heating with 3 N. hydrochloric acid at 150" C. 0. E. M.574 ABSTRACTS OF CHEMICAL PAPERS Rapid Estimation of Pyridine in the Presence of Ammonia in Gas-Water. 0. Baessler. (J. Gasbeleucht., 1912,55, 905-906 ; through Chem.Zentralbl., 1912, II., 1406.)-In the following modification of Houghton’s method the ammonia and pyridine are first estimated together by titration with or & sulphuric acid. Another sample of the gas-water is then heated, and the vapours conducted through two slightly-warmed aqueous solutions of sodium hypobromite, whereby the ammonia is oxidised to nitrogen and water, whilst the pyridine passes on and is received in & sulphuric acid, the excess of which is titrated with standard alkali.The estimation may also be made by distilling the ammonia and pyridine, and adding to the distillate, after titration, sufficient sodium hydroxide solution to render the liquid slightly alkaline. I t is then used for the estimation of the pyridine.C. A. M. Direct Determination of Rubber. C. R. Boggs. (Eighth Int. Coy. APP. Chew., 1912, vol. 9, 45-58.)-The author describes a modification of Budde’s method for the estimation of rubber in vulcanised or unvulcanised samples. The pro- cedure is as follows: The sample is ground until it passes through a %)-mesh sieve; it is thoroughly extracted with acetone and the residue is dried in a reducing atmosphere, care being taken to avoid undue exposure to the air.An amount of the residue corresponding to about 0.1 grm. of pure rubber is placed in a tall shaped beaker of 300 C.C. capacity, together with 20 C.C. of toluene ; the beaker is covered and the toluene boiled gently for three hours. Most of the solvent is then evaporated off on the water-bath, 50 C.C.of carbon tetrachloride are added, cooled, and then 50 C.C. of Budde’s brominating mixture are introduced (cf. ANALYST, 1910, 35. 267). The covered beaker is allowed to remain for twenty-four hours at the ordinary temperature with frequent agitation. Fifty C.C. of absolute alcohol are then stirred in, the precipitate is allowed to settle over night, the solution is decanted through a paper filter free from chlorine, and the residue is washed first with a mixture of 2 parts of carbon tetrachloride and 1 part of absolute alcohol, and then with alcohol alone.The residue is also washed with boiling water by decantation until free from bromides, and is then transferred to the filter. I t is washed with a little alcohol and allowed to dry in the air.The residue is then separated from the filter and ground to powder if necessary. The filter-paper is torn up and burned in a porcelain crucible with a little fusion mixture at a low temperature. The powdered residue is then added and mixed with about 15 grms. of the fusion mixture. The contents of the crucible are slowly brought to fusion-point, and gently fused for about ten minutes.The melt is dissolved in water, the solution filtered, acidified with nitric acid, and the bromine is estimated volumetrically with silver nitrate and thiocyanate. The amount of rubber is caIculated from the bromine by the formula ‘*=0*425. No correction is applied for the sulphur in the bromide. With Br.4 purified Para rubber the method gives about 93 per cent.of pure rubber ; oxidised rubber is not estimated, so that the method may be applied to determine the tendency to oxidation of samples before and after exposure. J. F. B.ORGANIC ANALYSIS 575 Relationship between the Amount of Resins and the Viscosity of Rubber Solutions. J. G. Fol. (Eighth Int. Cong. App. Chem., 1912, vol. 9, 71-73.) -In making solutions of Castilloa plantation rubber, the author has observed that a low viscosity generally indicates a high percentage of resins, and vice versa.Either the resins decrease the viscosity of the rubber, or the rubber molecules themselves possess a low viscosity, when the material contains a large amount of resinous matter. The viscosities of 0-25, 0.5, and 1.0 per cent. solutions of the rubber in pure benzene are observed in the Ostwald double-bulb viscosimeter, the rate of flow of the solvent itself being taken as unity.The general validity of the above rule is shown by tables, although in certain exceptional cases other factors must come into play. After removal of the resins by boiling acetone, the viscosity of the purified rubber was in some cases higher, and in other case8 lower, than that of the original sample.The heating of the rubber, without de-resination, at the same temperature and for the same time as are employed for removing the resins, caused in all cases a considerable decrease in the viscosity of the solutions. J. F. B. Analysis and Testing of Rubberised Fabrics. M. B. getllk and J. gofka. (Eighth Int. Cong. App. Chem., 1912, vol.9, 89-93.)-1n the examination of a rubberised fabric, the cohesion, adhesion, elasticity, and resistance to rubbing and folding of the coating are first tested empirically. The cloth is then supported on a frame and water is poured on it. The degree of resistance to water is tested for several days both in the case of the original fabric and of the fabric after rubbing between the hands.The action of 5 per cent. ammonia solution and 5 per cent. acetic acid is noted as representing the behaviour of the cloth towards the liquids of the body. The weight per square metre, thickness, and tensile qualities are observed. A portion of the cloth is boiled with chloroform for three hours, and the condition of the coating is then tested by scraping. If it is swollen and comes away in large patches it contains at least 50 per cent.of caoutchouc. The separated coating is extracted with acetone, which removes resins, oils, wax, etc. The residue is saponified with alcoholic potash, which removes rubber subsiitutes ; the remainder is caoutchouc and mineral matters. The presence of a large amount of mineral matter or rubber substitute impedes the swelling with chloroform; coatings rich in resin swell most rapidly under the action of carbon bisulphide.The fabric, freed from its coating, is washed with hot alcohol, dried and weighed, then tested for its tensile properties with the Schopper dynamometer. Fabrics with a 40 per cent. coating are quite satisfactory, even if half the coating is composed of substitutes. Zinc oxide, magnesia, and chalk are not objectionable in reasonable quantities. J.F. B. Expanded Meissl-Hiller Table for Invert Sugar. E. W. Rice. (Eighth Int. Cong. App. Chem., 1912, vol. 8, 47-51.)-The table is intended for use in laboratories in connection with the control of factory working where determina- tions are required of invert sugar in many samples of widely varying composition. AS the total amount of all sugars present, as well as the ratio between cane sugar and invert sugar, affects the precipitation of copper, the calculation of576 ABSTHACTS OF CHEMICAL PAPERS results includes factors to cover the conditions in each individual analysis.These calculations are simplified by the use of these tables, which are constructed for the following conditions : Cuprous oxide is precipitated from 50 C.C.of mixed Fehling’s solution by 50 C.C. of sugar solution, and the equivalent of invert sugar is calculated according to the Meissl-Hiller method. The columns of the tables contain, first, equivalent weights of copper, cuprous oxide and cupric oxide, which represent the analytical determinations made. Parallel with these are columns showing the corresponding percentages of invert sugar for quantities of 1, 2, 5, 10, and 20 grms.of the sample per 100 C.C. of the solution. These columns are each further sub- divided according to the polarisation of the sample. It is essential that all the experimental conditions be maintained rigidly constant. I t was found, for instance, that different beakers, apparently identical, gave variations of as much as 0-5 per cent.of the copper precipitated, owing to differing thickness and bottom area ; variations between duplicate determinations, due to other causes than the beakers, were found to run about 0.36 per cent, in samples of high purity, and 0.24 per cent. in samples of low purity. J. F. B. Electrolytic Apparatus for Use in Invert Sugar Determinations.(Eighth Int. Cong. App. Chem., 1912, vol. 8, 75-77.)-The author B. B. Ross. 9 describes a piece of apparatus (see Fig.) which very materially simplifies the electrolytic determination of the cuprous oxide obtained in the estimation of cupric- reducing sugars, since the electrolysis is effected without removing the oxide from the filter. The filtering funnel illustrated has a stopcock in its stem, and a spiral of platinum wire serving as the anode is fused into the lower portion of the funnel.An asbestos filter-bed is deposited in this portion of the funnel, and the cuprous oxide is precipitated and filtered by the aid of the suction-pump in the usual manner. After the oxide has been washed with hot water, the stopcock is closed and the funnel is filled about three-quarters full of dilute nitric acid, containing about 4 C.C.of acid (1.42 sp. gr.} per 100 C.C. A platinum cylinder, as shown in the illustration, is then immersed in the acid to serve as cathode, and electrolysis is begun. A current, yielding not more than 1 C.C. of gas per minute, should be used to secure the most satisfactory deposition of the copper. Complete sohtion and deposition of the cuprous oxide as metallic copper are thus readily attained.J. F. B. Inversion of Sucrose Solutions by Means of Ammonium Chloride. F. Strohmer and 0. Fallada. (Eighth Int. Cong. App. Chem., 1912, vol. 8, 85-92.)-The nitrogenous constituents which are not easily removed from impure sugar products are known to affect the accuracy of the estimation of sucrose by theORGANIC ANALYSIS 577 Clerget inversion method, because their optical rotation in acid liquids is different from that in neutral solution.For this reason an inverting agent which has a neutral reaction at the polarisation temperature was desirable. Ammonium chloride reacts at high temperatures as a dissociated mixture of ammonia and hydrochloric acid, and therefore inverts cane sugar.The authors' attempts to use this salt in quantitative sugar analysis have, however, led to no successful result on account of the incompleteness of the inversion under practicable conditions of working. Fifty C.C. of the normal sucrose solution, when heated for one hour at 110' C. with 30 C.C. of a solution containing 5 grms. of ammonium chloride, showed a polarisation of +3.6, as compared with an original polarisation of 49.9.The employment of zinc dust or platinum black as catalytic agents produced no more satisfactory result. Complete inversion was obtained by heating at 110" C. for forty-five minutes, 50 C.C. of a half normal sucrose solution with 30 C.C. of saturated ammonium chloride solution in presence of a trace of acid.On standing in the cold for twenty-four hours this solution gradually attained a maximum laevo-rotation exceeding that of pure invert sugar, accompanied by a coloration indicating the destruction of a certain small amount of dextrose. Further experiments seemed to indicate that the desired end might be attained by neutralising the ordinary acid inverted solution with ammonia or sodium hydroxide, J.F. B. Detection and Differentiation of Vegetable Tannins. E. Stiasny. (Collegium, 1912, 483-499 ; through Chem. Zentralbl., 1912, II., 1405-1406.)-For the detection of pyrogallol tannins in the presence of catechol tannins, 50 C.C. of the solution (as used for analysis) are boiled for thirty minutes beneath a reflux con- denser with 5 C.C. of strong hydrochloric acid and 10 C.C.of 40 per cent. formaldehyde solution, and then cooled and filtered. An addition of 1 C.C. of 1 per cent. ferrous ammonium sulphate solution and 5 grms. of solid sodium acetate is now made, without shaking, to about 10 C.C. of the filtrate, and note taken whether a blue or violet coloration is produced at the bottom of the tube. Bennett's test may also be used, according to which 2 to 3 C.C.of the tannin solution are mixed with 2 to 3 C.C. of 10 per cent. sodium sulphite solution and 1 or 2 drops of 10 per cent. potassium ehromate solution. All catechol tannins give a green coloration, whilst of pyrogallol tannins, myrobalans, sumach, and gallotannic acid, give a bluish-red coloration, rapidly changing to brown ; and valonia, chestnut, and oak-bark extracts givo a deep reddish-violet fairly stable coloration.Schell's test for the detection of mangrove extract in natural and sulphited quebracho extract consists in boiling and rapidly cooling 20 C.C. of the tannin solution, covering it with petroleum spirit and mixing it with 1 C.C. of 20 per cent. cobalt chloride solution and 1 C.C. of strong ammonia solution.Quebracho extracts give a grey-green, mangrove extracts a brown-violet, mimosa a grey-blue, and algarobilla a light brown, coloration in this test. It is not always possible, however, to detect 10 to 20 per cent. of mangrove extract with certainty in this way. In Hoppenstedt's test 25 C.C. of the tannin solution are gradually mixed with 25 C.C. of a 1 per cent. quinine hydrochloride solution and filtered, and 5 C.C. of the clear filtrate of non-tannin constituents are treated successively with 1 C.C.of concentrated acetic acid, 2 c.c of acetone, and 5 C.C. of578 ABSTRACTS OF CHEMICAL PAPERS ethyl acetate, and shaken. In the presence of mangrove extract the lower layer will be dark yellowish-brown, but colourless in the case of other tannins. In the author’s experience certain quebracho extracts and wood pulps may also yield a faint yellow coloration in the lower layer even in the absence of mangrove extract.C. A. M. Tests to determine the Commercial Value of Wood Preservatives H. F. Weiss. (Eighth Int. Cong. App. Chem., 1912, vol. 13, 279-300.)-From 40 to 90 per cent. of the total cost of treating wood is in general due to the preservative; and as in ordinary practice it is usual to inject about 10 pounds of the oil per cubic foot of wood, although about + pound will prevent fungoid growth, it is obvious there is a wide margin for safe reduction.The present paper exhaustively summarises the conclusions arrived at as a result of an experimental investigation of the following properties : (1) The important chemical and physical properties of the preservative; (2) the effect of the preservative on the strength of the wood; (3) the ability of the preservative to penetrate snd diffuse through the wood ; (4) the per- manency of the preservative after its injection, including a study of its volatility and capability of being leached; (5) the combustibility of the treated wood; (6) the toxic efficiency of the preservative in inhibiting the grawth of wood-destroying fungi; (7) the corrosive action of the preservative on steel ; (8) the effect of the preservative on paint applied to the wood subsequent to treatment.A description is given of the methods by which each of the above points was investigated, accompanied by illustrations of the special apparatus and plant required for many of them. For the results which follow, a greater accuracy than plus or minus 10 per cent. cannot be claimed, chiefly because of the variables in- herent in wood. The results themselves are too numerous for adequate abstraction, but a few of the general conclusions follow. The depth to which oils can be impreg- nated varies inversely as some function of their viscosity. To secure the best results, the wood and the preservative should be heated to the optimum temperature during the pressure period. With other than water-soluble preservatives the treatment should not be made too rapidly. With coal-tar creosote the fractions of greatest stability are the least toxic. Present practice favours the retention in treated wood of the more volatile fractions by an admixture of the more stable ones. It would appear that in practice there is being forced into wood about two and a half times as much zinc chloride, and fifty times as much coal-tar creosote, 8s is necessary to prevent decay; more thorough diffusion of less preservative rather than saturation of the outer layers of the wood is thus required. A prolonged seasoning of treated wood considerably raises its ignition temperature. Wood treated with the water-soluble salts mentioned in the tests, although less difficult to ignite, yet proved to be slower burning and more readily extinguishable than untreated wood. All the metallic salts used were much more corrosive in their action on steel than coal-tar creosote. All the oils tested rendered the wood unfit for subsequent painting, copperised oil being the least objectionable. H, F. E. H.

ISSN:0003-2654    

DOI:10.1039/AN9123700565

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

INORGANIC ANALYSIS 579 INORGANIC ANALYSIS. Estimation of Arsenic in Organic Compounds. T. St. Warunis. (Chem. Zeit. 1912 36 1205.)-For the analysis of organic arsenic compounds the combustion of which is difficult from 0.2 to 0-4 grm. of the substance is stirred with a platinum wire into an intimate mixture of 10 grms. of potassium nitrate and 5 grms. of sodium peroxide in a large nickel crucible covered with a layer of the mixture and heated gradually to fusion. When quite fluid the melt is stirred maintained for a short time in a state of fusion and dissolved in hot water. The solution which contains all the arsenic as arsenate is acidulated with hydrochloric acid filtered if necessary, and neutralised with ammonia. The arsenic is determined as Mg,As,07 by means of magnesia mixture.I t may also be estimated volumetrically as uranium arsenate (Briigelmann Zeit. anal. Chem. 1877 16 20) in the dissolved melt. The acidulated liquid is first rendered alkaline to litmus-paper with sodium hydroxide or ammonia, and then strongly acid with acetic acid and made up to 100 C.C. ; not more than 50 C.C. of this is taken for titration with a uranium solution containing about 20 grms. uranium oxide per litre. Non-volatile compounds which are readily decomposed are fused with potassium nitrate and sodium peroxide in a hard glass tube 35 cm. long, closed at one end. This is charged in succession with a 2-cm. layer of sodium peroxide the mixture containing the substance 3 cm. of sodium peroxide a plug of asbestos and anhydrous sodium carbonate for the rest of the tube.The sodium carbonate ie cautiously heated first then the sodium peroxide on each side of the substance and finally the whole tube the cooled contents of which are dissolved in water and treated as above. 0. E. M. Determination of Small Quantities of Arsenic by the Gutzeit Test. W. S. Allen and R. M. Palmer. (Eighth Int. Cong. App. Chem. 1912 VOI. 1, 9-17,)- Under standard conditions the Gutzeit test gives results as accurate as those obtainable by the Marsh-Berzelius method which requires much more of the analyst's time. The influence of small amounts of iron is recognised and met by the addition of from 0.05 to 0.1 grm. of iron as ferric salt in every test. The ferric salt i s reduced by stannous chloride and the presence of both iron and tin insures an even and adequate evolution of hydrogen and longer stains are obtained in this way than when omitting the addition of iron.The apparatus used by the authors consists of a 60-C.C. bottle the rubber stopper of which has a single hole. Through this hole passes the tapered end of a tube 7 cm. long and 1.2 cm. in diameter. This tube holds a strip of dry lead acetate paper and its rubber stopper carries a similar but shorter (4 cm.) tube which contains glass-wool moistened with a solution of lead acetate. Through the rubber stopper of this upper tube passes a tube 10 cm. long 4 mm. in internal diameter, and constricted in the middle to suwort in its upper half the strip of mercuric chloride paper. This is made by dipping filter-paper in 0-5 per cent.mercuric chloride solution removing excess of Bolution by use of a rubber roller and drying rapidly on a cloth in an oven at 100' C. The edges of the paper are trimmed of 580 ABSTRACTS OF CHEMICAL PAPERS and rejected and the rest is cut into strips about 7 cm. long and as nearly as possible 4 mm. wide. The reagents required are (1) zinc in pieces which pass a screen of three holes to the linear inch but not one of six holes to the inch. This zinc is washed in hydrochloric acid (1 1) and preserved under water. (2) Sulphuric acid (1 4), containing 100 grms. of sodium chloride in each litre the reduction of iron proceed-ing more readily in presence of hydrochloric acid. (3) Stannous chloride solution, made by dissolving 80 grms. of the crystallised salt in 100 C.C.of water and 5 C.C. of hydrochloric acid. The solution to be analysed which must not contain more than traces of sulphide or sulphite nor nitric acid chlorine etc. has iron alum solution added to it in amount corresponding to about 0.05 to 0.1 grm. of ferric oxide. This is followed by 0.5 C.C. of stannous chloride solution and the mixture is boiled until the iron is all reduced to the ferrous state when it is cooled and transferred to the Gutzeit apparatus together with fifteen pieces of zinc and 10 C.C. of the sulphuric acid mixture. After gently shaking its contents the bottle is placed in a water-bath at a temperature within a few degrees of 75O C. and after thirty minutes the sensitised paper is withdrawn dipped in melted paraffin and compared with a series of standards ranging from 0.001 to 0.02 mgrm.Standards preserved in sealed tubes over phosphoric anhydride and kept in the dark show no perceptible change in several months. G. C. J. (4) A solution of iron alum. Titration of Arsenic Acid. A. W. C. Menzies and P. D. Potter. (Eighth h t . Cony. App. Chern. 1912 vol. 1 367-371.)-When arsenic acid is titrated with sodium hydroxide using phenolphthalein as indicator a rather uncertain end point is reached when about two-thirds of the hydrogen has been replaced by sodium. Quite sharp end points corresponding to the formation of disodium arsenate may be secured as follows. To the solution to be titrated 15 C.C. of a oold saturated solution of barium chloride is added and the mixture is boiled to expel carbon dioxide.I t is then titrated with caustic alkali with stirring after each addition until the locally formed precipitate becomes slow in redissolving. The walls of the vessel are now if necessary scratched below the surface of the clear solution until the liquid on stirring appears lustrous with fine crystals after which the titration is completed. Under the above conditions the results are as accurate as the results of a volumetric analysis can be but the scratching action of a stirring rod to induce the formation of a precipitate of dibarium arsenate is necessary or the proper end-point may be much overstepped. In fact with solutions of known titer it is possible to run in alkali until practically the whole of the arsenic exists as trirtrsenate before the solution assumes a pink tint.G. C. J. Volatility of Arsenious Chloride. J. I. D. Hinds. (Bighth I&. Coy. App. Chem. 1912 vol. 1 227-231.)-Experiments are described which show that the quantity of arsenious chloride lost on boiling its hydrochloric acid solution is a, €unction of the concentration of the arsenic and the acid and that the rate of Ioss is muoh smaller than is supposed by some authors who issue warnings against boilin INORGANIC ANALYSIS 581 acid arsenious solutions even in qualitative work. arsenious chloride, of more than one-thousandth part of the arsenic present. For example a TG solution of as regards acidity may be concentrated to half-bulk without loss G. C. J. Detection of Traces of Arsenic and Antimony. D. R. Staddon. (Chem. News 1912 106 199.)-From 0.5 to 2 grms.of the substance to be examined are dis-solved in about 5 c . ~ . of water. Acids and acid salts are first neutralised. An equal weight of sodium hydrosulphite ( 6 ‘ blankit ’ I ) is added to the solution and the liquid warmed. Rapid reduction to metallic arsenic or antimony as the case may be occurs, the precipitate varying from a light to dark brown colour according to the quantity of substance present. The precipitate of arsenic is dissolved by sodium hypochlorite, while that of antimony is insoluble and remains suspended. One part of sodium arsenite in 50,000 parts of water may readily be detected by this method. A. R. T. Gravimetrie Estimation of Beryllium. B. Bleyer and K. Boshart. (Zeitsch. anaZ. Chem. 1912 51 748-755.)-The precipitation of beryllium as hydroxide is best effected in the cold using but a small excess of ammonia and taking care to have ammonium chloride present.Large amounts of the latter introduce no error but 1 grm. is usually more than sufficient. The precipitate is washed witb a hot dilute solution of ammonium nitrate containing a few drops of ammonia and any precipitate which adheres to the precipitating vessel is removed with a fragment of ash-free paper. The filter-paper and precipitate are transferred to a platinum crucible without previous drying and when the paper is burnt the crucible is covered and heated in the blowpipe flame for ten minutes. The precipitation should take place in a vessel of porcelain and not in a glass beaker. Platinum serves as well as but no better than porcelain.Precipitation in hot solution followed by boiling to eliminate excess of ammonia as is often recommended may lead to beryllium being under-estimated by as much as 20 per cent. Precipitation by means of ammonium sulphide (cf. Crookes’ t 6 Select Methods,” 1894 150) instead of ammonia gives equally good results but only if the precipitation be made in the cold and a considerable time allowed to elapse before filtration. Glassmann’s method (Ber. 1906 39 33683, depending on the use of a mixture of potassium iodide and iodate gives accurate results but occupies more time than the ammonia method. G. C. J. Sensitive Reaction of Free Bromine. G. Deniges. (Compt. rend. 1912, 155 721-723).-An aqueous solution of rosaniline bisulphite gives a purple or violet coloration and eventually a violet precipitate with free bromine.The reagent is prepared by adding 10 C.C. of sodium bisulphite solution (sp. gr. 1-29) to 1 C.C. of a 0.1 per cent. aqueous solution of fuchsin followed after about five minutes by 20 C.C. of hydrochloric acid (sp. gr. 1.18). The compound of rosaniline and bromine is insoluble in water ether benzene and petroleum spirit ; sparingly soluble in rosaniline bisulphite solution but very soluble in methyl ethyl and amyl alcohols, acetone acetic acid and carbon tetrachloride. The solutions of the compound show a characteristic spectrum with two absorption bands one in the blue and the othe 582 ABSTRACTS OF CHEMICAL PAPERS in the orange part of the spectrum. The action of bromine in the bisulphite is inhibited by adding hydrogen peroxide.In applying the test from one drop to several C.C. of the solution under examination are shaken with a mixture of 2 C.C. of the reagent 2 C.C. of hydrogen peroxide solution and 1 C.C. of chloroform and in the presence of as little as as 0.01 mgrm. of bromine a violet tint will be produced, Atmospheric air is tested by being made to bubble through the mixture or test-papers prepared by saturating filter-paper with the reagent and drying it in the air, may be used. Before use the paper is slightly moistened with water containing 1 to 2 per cent. of hydrochloric acid. The test may be used in the presence of a, large excess of chlorides or iodides and may be made quantitative by colorimetric comparison with standard solutions of bromine.The presence of combined bromine may be detected e.g. in 2 C.C. of a solution of potassium bromide corresponding to 0013 mgrm. of bromine by adding 0.5 C.C. of a 10 per cent. solution of potassium chromate and 0.5 C.C. of sulphuric acid shaking the mixture with a fragment of marble to cause a slow evolution of gas and suspending a moistened and acidulated test-paper 2 to 3 cm. above the surface of the liquid. C. A. M. Estimation of Free Carbon Dioxide in Water by Titration with Alkalis in the Presence of Phenolphthalein. J. Tillmans and 0 Heublein. (Zeitsch. Untersuch. Nahtr. Genussm. 1912 24 429-449.)-The authors show by their experiments that free carbon dioxide in water can be titrated acourately by dilute alkali solutions using phenolphthaleln as indicator.The quantity of carbon dioxide present in the water should not exceed 100 mgrms. per litre and the temporary hardness must not exceed 10 (German) degrees; otherwise the water must be diluted with neutralised distilled water previous to the titration. The quantity of phenolphthalein added is also of importance and it is recommended that 1 C.C. of a 0.035 per cent. alcoholic solution of the indicator be used for every 200 C.C. of water. The titration must be carried out in a closed flask (cf. ANALYST 1911,36 362). w. P. s. Use of Higher Phenols in Testing for Free Lime in Portland Cement. D. C. MeFarland and H. F. Hadley. (Eighth Int. Cong. App Chem. 1912, vol. 5 83-90.)-Experiments described in detail have shown that the salts formed by calcium with phenol and homologous compounds are only sparingly soluble in any of the ordinary solvents.The speed with which different phenols react with free lime shows great variations. For example combination takes place immediately in the case of p-cresol whilst P-naphthol and phenol do not react so rapidly and there is only a slight reaction with a-naphthol o-ethyl phenol carvacrol 1-3-4 xylenol, o-cresol and p-nitro-cresol. Of these phenols only P-naphthol and p-cresol appeared to react to the same extent as phenol. A rough estimation of the free lime in cement (when small in quantity) may be made by shaking 3 grms. of the sample for an hour with a mixture of 21 C.C. of absolute alcohol 30 drops of p-cresol and 0.5 C.C. of water then filtering the liquid washing the residue with water until colourless evaporating the filtrate and washings igniting the residue and weighing it as calcium oxide.If much free lime is present however the results thus obtaine INORGANIC ANALYSIS 583 are too low and do not agree well in duplicate estimations. A mixture of 1 C.C. of p-cresol 3 C.C. of absolute alcohol and 4 drops of water gives a reagent which reacts much more rapidly than White’s phenol reagent (ANALYST 1909 34 71) for the detection of free lime. For example an abundance of pale red crystals easily observed without polarised light was obtained in two to three minutes from cements containing 5 per cent. of free lime and by means of this test it was proved that in old cements free lime may still be present its particles having been protected by an outside coating formed by the action of the air.For practical purposes this modifi-cation of the test is regarded as almost too delicate. C. A. M. Estimation of Chlorous Acid (Chlorites). G. Lasegue. (BUZZ. SOC. Chim., 1912 11 884-886.)-The process is based on the insolubility of lead chlorite in alcohol ; this salt is slightly soluble in water but practically insoluble in 80 per cent. alcohol. I t slowly oxidises alcohol at the ordinary temperature but this action is very slight within a period of one hour. The chlorite is dissolved in water an excess of lead nitrate solution is added and the mixture is stirred. The mixture then receives the addition of six times its volume of 85 per cent. alcohol and after the lapse of one hour the precipitate is collected on a weighed filter dried over sulphuric acid under reduced pressure and weighed.When the chlorite solution contains an excess of alkali as is often the case when the chlorite has been prepared by passing the gas into an alkali solution magnesium nitrate is added before the estimation is carried out. The presence of small quantities of chlorides does not interfere with the estimation and chlorites chlorides hypochlorides and chlorates may be estimated in a mixture containing the same in the following way In an aliquof portion of the solution the chlorite is estimated as described. Another portion of the solution is treated with potassium iodide and hydrochloric acid the iodine liberated being ti measure of the chlorite plus hypochlorite. In order to prevent the action of the chlorate on the iodide the solution must be very dilute and contain not more than 0.3 grm.of the salt per 500 C.C. In a third portion of the solution the chloride, resulting from the reduction of chlorite and hypochlorite by a standardised alkali arsenite solution together with the chloride originally present is titrated with silver nitrate solution. After removing the silver chloride by filtration the chlorate in the filtrate is reduced by means of sodium nitrite and the resulting chloride also titrated. The following results were obtained with a mixture of the four salts the figures in parentheses showing the actual quantities present Chlorine as chloride, 13.1 (12.5) j as hypochlorite 11-92 (12-5) ; as chlorate 37.48 (37.5) ; as chlorite, 37.34 (37*5) w.P. s. Quantitative Separation and Estimation of Copper by Means of Hydroxylamine Hydrochloride. A. Bayer. (Zeitscht. anal. Chem. 1912 51, 729-735.)-To the solution of copper as sulphate which may conveniently contain about 0.3 grm. copper in 50 c.c. there is added about 50 C.C. of 25 per cent. solution of Rochelle salt 25 C.C. of a 15 per cent. solution of sodium hydroxide and 2 C.C. of 8 5 per cent. solution of hydroxylamine hydrochloride. The mixture is boiled for a minute and the cuprous oxide is filtered off in a tared Soxhlet tube ignited to cupri 584 ABSTRACTS OF CHEMICAL PAPERS oxide and weighed. Neither antimony zinc bismuth lead iron arsenic nor tin interferes and quantities of ammonia or of nitrates corresponding to 1 grm. of ammonium chloride or sodium nitrate do not influence the results which are exact.Mercury if present is partly precipitated but introduces no error if the cuprous oxide is ignited to cupric oxide. Silver gold platinum manganese cobalt and nickel should be absent. If the three latter metals but not gold platinum or silver, be present they are easily eliminated by precipitating the copper first as sulphide in acid solution. The precipitate of copper and other sulphides together with the filter-paper is oxidised by nitric acid and the solution taken down to fuming with sulphuric acid then diluted and copper precipitated as described. G. C. J. Electro-analysis of Copper Antimony Bismuth and Tin with Acidified Chloride Electrolytes. E. P. Schoch and D. J. Brown. (Eighth Int. Con# App.Chem. 1912 vol. 21 81-91.)-Copper tin antimony and bismuth can be deposited quantitatively and fairly quickly from hydrochloric acid solutions provided a reducing agent such as hydroxylamine hydrochloride is present. An alloy of copper and tin may be analysed as follows The alloy is dissolved in aqua regia and the solution evaporated to expel free chlorine and oxides of nitrogen. Hydroxylamine hydrochloride (2 grms.) and about 10 C.C. of concentrated hydrochloric acid are then added and the solution is diluted to 200 C.C. and electrolysed using the Sand-Fischer apparatus (Zeitsch. Electrochem. 1907 13 469) and limiting the cathode potential to -0.4 volt (against a normal calomel electrode). The copper is deposited in from twenty to forty minutes and the end-point is recognised by the fact that the current must be reduced to zero to prevent the cathode potential exceeding the limiting value.The beaker is lowered and the cathode is washed detached dipped in alcohol and ether dried above a flame and weighed. It is then replaced 3 grms. of hydroxylamine hydrochloride is added to the electrolyte and the tin is deposited by means of a constant current of 1.5 amps. the end point which is reached in twenty to forty minutes being ascertained by means of hydrogen sulphide. With more than 0.7 grm. tin present in 200 c.c. a lower current density is necessary to ensure good deposits. In the electrodeposition of antimony the evolution of hydrogen and attendant stibine must be avoided the temperature must be above 50’ C. or “explosive” antimony may be deposited and unnecessarily high temperatures must be avoided, as they increase the tendency of antimony solutions to hydrolyse.Good results are obtained at 50” to 7 5 O C. with 20 to 25 C.C. of hydrochloric acid and 2 grms. hydroxylamine hydrochloride in each 200 C.C. of electrolyte and a cathode potential of -0.4 volt (against a normal calomel electrode) ; 0-5 grm. may be deposited in fen minutes with an error not exceeding 1 mgrm. To obtain a good deposit of bismuth it is necessary to limit the cathode potential to -0.25 volt until most of the metal has been deposited and then to allow it to rise gradually to -0.6 volt during the last few minutes. At 55’ to 75” C. with 5 C.C. hydrochloric acid and 2 grms. hydroxylamine hydrochloride present in 200 C.C.of electrolyte 0.5 grm. of bismuth may be deposited in twelve minutes with an error not exceeding 0-8 mgrm. G. C. J INORGANIC ANALYSIS 585 Influence of Lead on the Ferrocyanide Titration of Zinc. V. Lenher and C. C. Meloche. (Eighth Int. Cong. App. Chem. 1912 vol. 1 279-284.)-Con-trary to the statements of several authors (e.g. Seaman J. Amer. Chem. Soc. 1909,. 29 207) lead does not interfere with the estimation of zinc by ferrocyanide as that operation is usually carried out-namely in presenca of hydrochloric acid and am-monium chloride. The concentration of the latter may be anything between 1 and 20 grms. in 200 c.c. but concentrations as high as 20 per cent. give rise to an indistinct. end-point. Indistinct end-points also result from too low or too high concentration of acid but sharp end-points are obtained over a fairly wide range-namely from 1.5 to 6 per cent.free hydrochloric acid. Ten C.C. of concentrated hydrochloric acid and 10 grms. of ammonium chloride in 200 C.C. are convenient quantities to have present. The best indicator for general use is a 5 per cent. solution of uranium nitrate. A 1 per cent. solution of ammonium molybdate is more sensitive but its application is limited. G. C. J. Rapid Determination of Magnesia in Limestone by Means of the. Hydrogen Electrode. (Eighth Int. Cong. App. Chern. 1912 vol. 1,217-225.)-The method depends on the fact that when caustic alkali is added to a neutral solution containing salts of calcium and magnesium, the concentration of hydroxyl ions increases very slowly until all the magnesium is-precipitated as hydroxide and then increases rapidly to the point at which calcium hydroxide begins to separate.The method therefore consists essentially of the titra-tion of the magnesium making use of the above-described sharp rise in hydroxyl ion concentration to determine the end-point. In practice it is more convenient to measure the diminishing hydrion concentration which is a reciprocal function of the hydroxyl ion concentration. The necessary apparatus includes an easily constructed form of hydrogen electrode a calomel electrode a voltmeter reading from 0.01 to 1 volt any fairly constant cell with E.M.F. not less than 1 volt a sliding rheostat or bridge wire of fairly high resistance and a galvanometer or Lipmann electrometer as zero indicator.The limestone is dissolved in hydrochloric acid and the solution is freed from carbon dioxide and titrated with normal caustic alkali (free from carbonate) in a beaker into which dip the hydrogen electrode and the syphon from the calomel electrode. By the use of methyl-orange the neutralisation of the free acid may be quickly effected. From this point onwards electrometric readings are taken and the observed E.M.F. in centi-volts plotted as ordinates against the burette readings as abscissae. With a small addition of alkali beyond that necessary to change the colour of methyl-orange the E.M.F. rises rapidly and a further small addition causes the curve to take a nearly horizontal direction. Larger additions of alkali may now be made between each reading until the curve assumes a steeper gradient when readings are taken at every 0.2 C.C.When the gradient of the curve begins to diminish instead of increasing the titration is at an end. The reading of the burette ati this moment less the reading when curve changed its nearly vertical direction to a horizontal one is the measure of the magnesium present. The method is accurate to within 2 per cent. of the magnesium present. The paper includes a diagram of the whole apparatus a drawing of the hydrogen electrode and a typical curve. J. H. Hildebrand and H. S. Harned. G. C . J 586 ABSTRACTS OF CHEMICAL PAPERS New Volumetric Method for the Estimation of Nitrites and for the Separation of Nitrous Acid from Nitric Acid. W. M. Fischer and N.Stein-bach. (Zeitsch. anorg. Chem. 1912 78 134-140.)-The method depends on the readiness with which nitrous acid is converted into its methyl ester (b.-pt. - 12O C.), I t is sufficient to add a nitrite solution drop by drop to a mixture of dilute sulphuric acid methyl alcohol and water through which a current of air is bubbled in order to convert the whole of the nitrite into the ester and remove the latter as it is formed. If a known quantity of sulphuric or hydrochlorio acid is present in the mixture the difference between this and the amount remaining at the end of the experiment is a measure of the nitrous acid. None of the nitrite is oxidised to nitrate during the reaction and the mixture does not yield a reaction for nitrate at the end of the operation unless nitrates were present originally in the nitrite solution.The estimation is carried out most con-veniently by placing the nitrite solution and 45 C.C. of methyl alcohol in a 300 C.C. flask. The flask is closed with a rubber stopper through which pass the stem of a tapped funnel an inlet tube reaching to the bottom of the flask and a delivery tube the latter being attached to an exhaust pump. A measured excess of & hydrochloric or sulphuric acid and 1 C.C. of methyl alcohol are placed in the funnel and allowed to drop into the flask while air is drawn through the apparatus. When all the acid has been introduced the current of air is passed through the flask for a further five minutes and the excess of acid is then titrated with Fa sodium hydroxide solution (free from carbonate) using phenolphthalein as indicator.If instead of a nitrite solution a mixture of nitrite and nitrate is under examination the nitrate may be estimated subsequently by adding to the neutralised contents of the flask 3 grms. of Devarda’s alloy and 20 grms. of potassium hydroxide and estimating the ammonia formed in the usual way. The method may also be employed for estimating nitrites and carbonates together as the latter will neutralise an equivalent quantity of the acid in the flask ; st separate estimation of the nitrous acid for instance by titration with permanganate will allow the amount of carbonate to be calculated. Results of analyses are given which show that the method is accurate. w. P. s. Determination of Oxygen in Iron and Steel by Reduction in an Electric Vacuum Furnace.W. H. Walker and W. A. Patrick. (Eighth Int. Cong. App. Chem. 1912 vol. 21 139-148.) - Ledebur’s method for the determination of oxygen in iron and steel gives low results because hydrogen at a red heat fails to reduce alumina and Bilica and the reduction of oxides of manganese is far from complete. The method now described depends on the reduction of all the oxides to carbides by heating them with excess of carbon to a high temperature in an electric furnace the oxygen being given off quantitatively as carbon monoxide. The furnace illustrated in the paper is a vacuum furnace of the Arsem type (made by the General Electric Company) and requires 14 kw. (200 amperes at 70 volts) to operate it. The charge of metal taken for the test is 20 to 25 grms.and is contained in a graphite crucible and covered with about 5 grms. of powdered graphite. The furnace is exhausted to 0.01 mrn. by means of rotary oil-pump working in series with a Geryk pump the temperature being raised to 500Oto 600’ C INORGANIC ANALYSIS 587 towards the end of the operation. The furnace is allowed to cool and nitrogen dried by means of sulphuric acid and phosphoric anhydride is admibted until the gas pressure is about half an atmosphere. The nitrogen is then exhausted by the pump heating to 500" to 600" C. when the pressure is reduced to 0.01 mm. The charge is now fused an operation which requires three or four minutes. A mica window enables the operator to watch the fusion and switch off the current if as is often the case there is a tendency to violent ebullition at the moment of complete fusion.After allowing the charge to cool slightly the circuit is completed once more and heating continued for twenty minutes. A portion of the gaseous contents of the furnace is now transferred to a glass container by means of a Tiipler pump, and is dried and passed over iodine pentoxide heated to 130" C. the liberated iodine being retained by a solution of potassium iodide and titrated with thiosulphate. The fraction of the gaseous contents of the furnace taken for the final test is determined by the readings of a differential gauge attached to the furnace. When operating as described the blank to be subtracted from the oxygen found lies between 0.012 and 0-013 grm. This is mainly due to traces of oxygen which obstinately resist removal during the preliminary exhaustion.One steel examined showed 0.10 per cent. oxygen by the new method and nil by Ledebur's; another 0.23 by the new method and 0.069 by Ledebur's. G. C. J. Determination of Ozone by Means of Cadmium Potassium Iodide. C. Baskerville and W. J. Crozier. (J. Amer. Chem. Xoc. 1912,34,1332-1337.)-Schonbein's method for the estimation of ozone by determining the iodine liberated from metallic iodides is quite generally employed although open to objections ; and the fact that when acid potassium iodide solution is used the reaction is quite unreliable is well known. The authors have confirmed and amplified the work of Basberville and Hamor and found that pure cadmium potassium iodide (CdKI,.H,O) is much more stable than potassium iodide and is a satisfactoryareagent for the estimation of ozone where neutral solutions of potassium iodide cannot be employed.The cadmium potassium iodide of the above composition must be pure free from cadmium iodide potassium iodide and excess of water. Such a reagent was found to be as delicate as potassium iodide solution of the same strength for hydrogen peroxide and ozone. In the estimations the air to be ozonised was passed from a gasometer through the ozoniser and a bulb (Meyer's) apparatus for holding the solution of the reagent this apparatus being connected by a ground-glass joint. Similar results were obtained with neutral and acid solutions of & strength of the cadmium reagent and neutral potassium iodide while the results with acid potassium iodide were higher.Satisfactory results were obtained by absorbing the ozone in turpentine oil liberating the iodine by agitating the oil with the iodide solution and titrating the iodine with sodium thiosulphate solution. The results with alkaline iodide solutions were unsatisfactory. Cadmium potassium iodide solution acidified with hydrochloric acid is more reliable than potassium iodide in neutral solution but its employment is only necessary when the ozone is absorbed in turpentine oil. A. R. T 588 ABSTRACTS OF CHEMICAL PAPERS Potash Analysis (Potassium-Sodium Cobaltinitrite Method). E. A. Mitscherlich and H. Fischer. (Landw. Vm.-Stat. 1912,78 75-86 ; through Chem. Zentralbl. 1912 II. 1312.)-The authors have made further investigations on the com-position of the precipitate of potassium sodium cobaltinitrite and on the potassium factor deduced therefrom (see also Landw.Vem-Stat. 1910 76 139). The composi-tion of the precipitate changes with the concentration of the reagent. With increasing proportions of the reagent (0.3 grm. of cobalt chloride + 0.5 grm. of sodium nitrite} the quantityof sodium entering the molecule of the product increases at the expense of the potassium. The composition of the salt approaches a limiting value asymp-totically after a certain excess of reagent has been added. If this excem amounts to thirty or more times as much sodium as potassium the variations in composition of the precipitate fall within the limits of error of the estimation. The method, therefore gives satisfactory results provided a large excess of precipitating reagent be taken ; if more than 30 parts of sodium to 1 of potassium be used it is immaterial how great the excess may be.The reduction factor to be used for calculating from ;G potassium permanganate to potash (KeO) is 0.000161 not 0*000157 as previously given (Zoc. cit.) . The authors have also shown that their method is applicable for the estimation of potash in fertilisers. With quantities of potash from 0.002 grm. of K20 upwards, the probable error is on the average less than k 1 per cent. of the quantity to be estimated. J. F. B. Estimation of Potassium in Potassium Silicate. E. Wilke-Dorfurt. (Zeitsch. anal. Chem. 1912 51 755-760.)-For the estimation of potassium in “phonolith,” a low grade potash manure in which the potash exists as silicate, Verweij (Zeitsch.anal. Chem. 1909 48 760) recommended fusion with calcium carbonate and ammonium chloride as in the ordinary Lawrence Smith method for the estimation of alkalis in silicates followed by the direct precipitation of the potassium in the aqueous extract from the fusion without previous separation of lime. The convenience of the Lawrence Smith method which gives the alkalis as chlorides is admitted as well as its accuracy but analytical numbers are now brought forward to show that it is necessary to remove from the solution the very large amounts of calcium chloride which result from the reaction in the crucible, before proceeding to precipitate potassium as platinichloride.If calcium salts are not removed 9 per cent. of potash may be reported when only 7 per cent. is present. Such differences have occurred between the results of different German chemists reporting on one and the same sample of “phonolith.” Fresenius and Brinton showed (ANALYST 1911 36 178) that potassium could be estimated with fair accuracy as platinichloride in presence of an equal weight of calcium but in the solution resulting from the fusion of phonolith with calcium carbonate and ammonium chloride the ratio of calcium to potassium is much greater than 1 1. G. C. J. Mechanical Analysis of Arid Soils. W. Beam. (Cairo Sci. J. 1912 5 107-119 ; through Bull. of the Bureau of Agric. Intell. and of Plant Diseases 1912,3,1721.) -Soils of tropical and arid regions differ from those of other districts iu thei INORGANIC ANALYSIS 589 calcium carbonate content and especially in the amount of calcium sulphate in the flubsoil and in the low humus content which in semi-arid regions may reach less than 0.5 per cent.Clay soils-i.e. those which contain about 50 per cent. of clay-yield only about half the clay in the form of a suspension when shaken in a rotating machine (100 motions per minute) unless a 2 per cent. solution of sodium carbonate is employed in place of water. A threefold division is recommended for grading soils of this type (a) particles of more than 2 mm. ; (b) particles lying between 2 mm. and 0.002 to be separated into particles of 0-2-2 mm. 0.02-0.2 mm. and 0.002-0*02 mm.; ( c ) particles of less than 0.002 mm.the clay. The following procedure is adopted : 100 grms. of air-dried soil are passed through a sieve with 2 mm. round holes. The fine soil is afterwards separated by a 1 mm. sieve and the sandy residue then washed. The fine portion is dried at 110' C. 5 grms. are weighed out treated with a 2 per cent. solution of sodium carbonate mixed by means of a cut-down camel-hair brush and transferred to a cylinder the volume being made up to 10 cm. The whole is allowed to stand for eight to twelve hours and the supernatant liquor is decanted off at intervals of eight hours being replaced by 0-5 grm. sodium carbonate and more water. After three decantations the residuum is again mixed up by means of the brush and after the removal of the clay is washed with distilled water in a 200 C.C.vessel and again left to settle for eight hours. The residue is then dried and weighed, the amount of clay being determined from the difference in weight the soluble salts and humus being deducted if necessary. H. F. E. H. Modification of the Method of Meehanical Soil Analysis. C. C. Fletcher and H. Bryan. (U.S. Dep. Agric. Bureau of Soils Bull. 1912 No. 48 ; through Bull. of the Buwau of Agric. Intell. and Plant Diseases 1912 3 1720.)-Two thousand soils yearly are analysed and added to the collection of 20,000 samples in the Bureau, most of which have been submitted to mechanical analysis. The following are the limits of mechanical grades Fine gravel 2-1 mm. ; coarse sand 1-0-5 mm. medium sand 0.5-0.25 mm. ; fine sand 0.25-0*1 mm. ; very fine sand 0*1-0.05 mm.; silt, 0*05-0*005 mm. ; clay 0*005-0.0009 mm. The modified method is as follows The field samples are passed through a 2 mm. sieve and the fine earth is then dried in aluminium dishes for two hours in an electric oven cooled in a desiccator and 5 grm. samples weighed out into 8-ounce bottles. To each bottle is added 2 ounces of water and a few C.C. of ammonia. All are then shaken in a shaking machine for at least seven hours. Each sample is then brought into suspension by the use of a compressed water-jet and the sands allowed to settle the time required being con-trolled by microscopic examination. The liquid portion is decanted into a tube and centrifuged until no silt particles remain in suspension. The clay suspension is poured into a porcelain vessel and the treatment repeated.Thus finally the sands are left in the bottles the silt in the centrifuge and the clay in the porcelain vessels. The sands are transferred to platinum dishes dried and their total weight obtained after which they are separated by three minutes' shaking in suitably meshed sieves. The silts are then transferred to platinum dishes dried and weighed and the clay water evaporated in enamelled-ware cups of 300 C.C. capacity the clay being weighed in them without further transference. The clay may also be determined by difference wit 590 ABSTRACTS OF CHEMICAL PAPERS an accuracy of 1 per cent. Detailed descriptions and drawings of all the apparatus are given. H. F. E. H, Action of Boiling Sulphurie Acid on Platinum. Le R. W. McCay.(Eighth Int. Cong. App. Chem. 1912 vol. 1 351-359.) - Boiling sulphuric acid dissolves platinum in considerable amounts. I t is shown that the oxygen of the air is not concerned in this action. In prolonged evaporations of sulphuric acid solutions in platinum basins the action of the acid on the platinum is almost entirely prevented by the presence of a globule of sulphur and it is suggested that any possible loss of platinum in the parting of silver-platinum alloys with boiling concentrated sulphuric acid could be avoided by a similar expedient. Note by Abstractor.-That platinum is lost in the ordinary platinum assay ha8 been shown by Steinmann (ANALYST 1911 36 605). G. C. J. New Colorimetric Method for Titanium. V. Lenher and W. 6. Crawford, (Eighth Int. Cong.App. Chem. 1912 vol. 1 285-293).-For the estimation of very small quantities of titanium a method depending on the use of a solution of thymd in sulphuric acid is preferable to the hydrogen peroxide method. The depth of colour produced which is strictly proportional to the amount of titanium present, is twenty-five times as intense as the colour produced by hydrogen peroxide. The substance to be tested is fused with potassium bisulphate and the fusion is taken up in concentrated sulphuric acid cooled and treated with the reagent. This is prepared by dissolving thymol in a little acetic acid and diluting to a convenient strength with concentrated sulphuric acid. The use of the acetic acid avoids the development of the yellow colour which is produced when thymol is dissolved directly in sulphuric acid.The ratio of thymol to titanium may vary within wide limits but it is found best t o have it not less than 100 1. If it is necessary to dilute the test solution or standard concentrated sulphuric acid should be used for the purpose as the colour fades appreciably when the concentration of sulphuric acid falls below 80 per cent. Fluorides bleach the colour. G. C. J. The reagent should not be exposed to direct sunlight. Volumetric Determination of Titanium. P. W. Shimer and E. B. Shimer. (Eighth Int. Coy. App. Chem. 1912 vol. 1,445-451.)-Reduction of titanic to titanous oxide by means of ordinary zinc is very slow. By the use of a reductor described in this paper the reduction may be accomplished in fifteen minutes. The reductor consists of a tube 3 feet long and Q inch internal diameter drawn out at its lower end, which is plugged with glass-wool.Above the glass-wool the tube is packed with amal-gamated zinc of a size that passes through a 10-mesh sieve and over 20-mesh. The drawn-out end of the reductor passes through one hole of the double-bored rubber stopper of a flask and reaches half-way to the bottom of the flask. To the upper end of the reductor a 4-inch funnel is attached. Before commencing a reduction the flask is filled with carbon dioxide and connected to the reducttor through which INORGANIC ANALYSIS 591 little hot dilute sulphuric acid has been passed to warm the zinc. The solution to be reduced is now placed in the funnel and drawn through the reductor slowly by applying moderate suction to the tube which pierces the second hole in the rubber stopper of the flask.The rate of flow should be such that the solution drops from the reductor and does not form a continuous stream. The reductor is washed by sucking through it hot water acidulated with sulphuric acid and the flask is finally disconnected ; any air contained in it is expelled by passing carbon dioxide through it and the solution is titrated with ferric solution using thiocyanate as indicator. In presence of a large amount of iron as in a solution of a steel it is necessary to put the solution twice through the reductor. In pig-iron the whole of the titanium is found in the residue insoluble in hydrochloric acid. Vanadium which interferes with the titration of titanium is most conveniently separated from the latter by fusion with sodium carbonate and sulphur.The aqueous extract of the melt contains all the vanadium but no titanium. The undissolved residue is then got into solution by appropriate means and the titanium determined as described. G. C. J. Fusion of Certain Rare Earths with Alkali Carbonates. Separation of Tungsten from Iron Beryllium and Aluminium. M. Wunder and A. Schapiro. (Ann. Chinz. anal. 1912 17 323-327.)-Fusion of oxides of cerium and lanthanum with sodium or potassium carbonates left insoluble residues in close agreement with theory provided that solid alkali carbonate was added at the time of boiling the mass with water. In the case of thorium oxide however the results were somewhat too low whilst with didymium and erbium oxides a large proportion remained in solution.For example on fusing 0.1940 grm. of didymium oxide for an hour with 4 grms. of sodium carbonate boiling the mass for twenty minutes with 200 C.C. of water to which 1 grm. of sodium carbonate had been added and filtering the hot liquid the insoluble residue weighed 0.1629 grm. whilst a crystalline deposit (0.028 grm.) separated as the filtrate cooled (loss-14*66 per cent.). The insoluble residue when fused again and treated as before yielded 0.1376 grm. of insoluble residue and 0.0251 grm. of deposit in the filtrate (loss = 154 per cent.) ; whilst after a third fusion the loss was 11 per cent. In corresponding tests with potassium car-bonate the respective losses of didymium oxide after the three fusions were 8-12 10, and 11.97 per cent.Erbium oxide fused for an hour with sodium carbonate and the mass boiled with water and alkali carbonate showed loss of 5.4 per cent.; whilst the losses after two successive fusions with potassium carbonate were 36 and 37.5 per cent. respectively. Separation OJ Tungsten.-On fusing tungstic oxide for fifteen minutes with about six times its weight of sodium carbonate or potassium carbonate the mass w a ~ completely soluble in water. For the separation of a mixture of tungstic and iron oxides the mixture (about 0.5 grm.) was fused for an hour with about 5 grms. of sodium carbonate the mass taken up with water and the liquid boiled for twenty minutes after the addition of 1 grm. of solid sodium carbonate. The undissolved iron was fused again in the acid and was finally dissolved in hydrochloric acid an 592 ABSTRACTS OF CHEMICAL PAPERS precipitated with ammonia. nitrate dissolved in very dilute nitric acid and neutralised by ammonia. The tnngsten in solution was precipitated by mercurous Taken. Grm. Found. Grm. first filtrate . 0.2950 second filtrate . . . 0.0023 first fusion . 0.1132 second fusion . 0.1102 Tungstic oxide . 0.2983 { Iron oxide . 0.1106 { Good results were also obtained in a single fusion continued for two hours. Tungsten and beryllium oxide were separated in exactly the same manner and with similar good results. In the case of aluminium oxide the mixture after fusion with sodium carbonate was treated with ammonium nitrate in excess to separate the two oxides. C. A. M

ISSN:0003-2654    

DOI:10.1039/AN9123700579

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

592 ABSTRACTS OF CHEMICAL PAPERS APPARATUS, ETC. New Calorimeter Bomb, with Speeial Advantages as to Material of Construction and Method of Operation. S. W. Parr. (Eighth Int. Coy. App. Chem., 1912, vol. 1, 389-393.)--The bomb is constructed from an alloy of nickel, copper, tungsten, and chromium, the composition and general properties of which are described in a paper communicated to another section (sec.2). The melting-point of the alloy is so high (1300" C.), and the shrinkage at the point of solidification so great, that sound castings are obtained with diEculty, but when obtained, they form an excellent substitute for platinum. Another author (Jesse, vol. i., 233-236) has shown that, in determining the heat of combustion of a sample of coal, not more than 0-2 mgrm. of copper and 0.8 mgrm.of nickel are dissolved by the combined nitric and sulphuric acids present in the products of combustion of coal. The heat of formation of the equivalent quantities of copper and nickel nitrate introduce an error of less than 0.02 per cent. into a, calorimetric determination. The other improvement described consists in the use of a rubber gasket in placo of the usual lead gaskeb, and the replacement of the ordinary steel needle valve by a seated valve bearing a rubber gasket.The portion of the gasket which comes in contact with the gases in the bomb is a ring about 0.001 inch in width, and to reach this the hot gases have to pass through a very narrow space, with massive metallic cooling surfaces on each side. There is only indirect proof that the use of rubber gaskets is without effect on the accuracy of the results obtained by means of the bomb.In two series of experiments with sugar and benzoic acid, the greatest deviation from the mean was 7 parts in 10,000; and the ratio of the heat of com- bustion of benzoic acid to that of sugar was found to be 1.6014, against 1.6003 found by Fischer and Wrede, and 1.6016 the value found by the Bureau of Standards.G, C. J. Electrically-Heated Microscope Slide. F. G. Cottrell. ( J . Amer. Chm. Soc., 1912, 34, 1328-1332.)-The apparatus has been found useful for the study of liquid crystals and for any observations under the microscope where an adjustment of temperature is desired. The slides are made by cutting, crosswise, +inch widthsAPPARATUS, ETC.593 from very thin (0-6 to 0.8 mm.) 1 x3 inch microscope slides, the cut edges being rubbed smooth on fine emery cloth or carborundum powder. Both ends of these little slides are then coated with gold or platinum, and the centre part is specially coated with a very thin transparent film of platinum, the platinised ends serving as terminals for the centre of the slide, and the film on the central portion acting as the electrical resistance for heating the slide.inch with gold porcelain paint, or dipped in a collodion solution to which a small quantity of gold or platinum chloride has been added, and the adhering film dried in the air. The organic matter is next; gradually burned off by means of a Bunsen flame, a brilliant metallic surface being left, which may be thickened if necessary by repeating the process.To obtain the far thinner film required for the centre of the slide, a stout test-tube, A (fig.), carrying a rubber stopper, is fitted with a tube, B, for connecting to the vacuum pump, an aluminium wire anode, C, and a glass tube, D, through which is sealed a stout platinum wire, E, carrying a welded stiff platinum plate, F The ends of the slide are painted for about (cathode).Electrical connection to the last-named is made by running mercury down the glass tube to G. The negative wire from the induction coil may be placed in this mercury. A mica plate, B, is slipped over the cathode tube to screen the end from deposit and enable the process to be observed as the platinising of the slide I proceeds.When the slide has been placed in position, the joints are sealed with melted beeswax containing 10 to 20 per cent. gutta-percha. The vessel is exhausted to 0.01 mm. of mercury, and this point is indicated by the green fluorescence of the glass when the current is applied. As soon as a slight but perceptible darkening of the central portion of the glass slide is observed, it is withdrawn.An ordinary glass slide (3 inches x 1 inch), fitted with binding posts and metal springs, forms a convenient holder for these platinised slides. The rheostat used with these slides consists of a glass cylinder (1 inch x 4 inches high) on a substantial foot. oork stopper pass two glass-stemmed platinum electrodes similar to the cathode in the figure, except that the platinum sheet may be omitted.One of the electrodes should reach to the bottom of the cylinder, the other sliding easily through the cork to allow adjustment of the distance between them. The cylinder is filled with distilled water, acidified with 1 drop of dilute sulphuric acid. For temperatures only slightly above room-temperature, only very faint acidity may be needed.In observing electrolytes, a thin cover-glass may be placed over the platinum surface, but in other cases the matetial is in actual contact with the slide. The temperature is easily controlled. Probably the temperature could be ascertained by measuring the current flowing through the platinum with a galvanometer while measuring drop in potential across the film by the compensation method, and from these data Through594 ABSTRACTS OF CHEMICAL PAPERS estimating the resistance of the film, which, in the case of pure platinum, should be directly proportional to the absolute temperature.A. R. T. Sodium Flame Lamps for Polarisation. E. Beckmann. (Ber., 1912, 45, 2523-2529.)-The objections to the ordinary form of polarisation lamp, in which solid sodium chloride is volatilised in a Bunsen flame, have led the author to design lamps in which the sodium coloration is obtained by the minute globules of liquid pro- jected from the surface of a solution of sodium hydroxide or carbonate which is undergoing electrolysis.Such a lamp is illus- trated in the figure. The body of the burner is constructed of porcelain or quartz-glass, and the gas is introduced through a removable side-tube having a small hole, r, which serves as a jet. By turning the jet more or less from the vertical direction the amount of air sucked in at the bottom may be regulated.The electrolyte contained in the vessel situated below the burner consists of a 2 to 4 per cent. solution of sodium hydroxide or carbonate, and plate electrodes of nickel are employed.Two accumulators in series or three dry cells provide the current, - and the evolution of gas must be so controlled that the liquid + appears merely milky, without the formation of large bubbles. If the electrolyte vessel be brought sufficiently close to the bottom of the burner-tube, the whole of the gas evolved is drawn in, and the intensity of the light is equal to that of the ordinary salt burner.The intensity of the flame may be increased by fitting a porcelain or iron shell around the top of the burner, as shown at 12, forming an annular orifice at the top, and closed at the bottom. A stream of oxygen, not under any appreciable pressure, is conducted through this annular shell. Another form of burner is also mAPPARATUS, ETC.595 described, in which the electrolytic bath surrounds the top of the burner, and only the outer zone of the flame is coloured by the sodium solution. When a flame of unusual intensity is required, a burner suitably modified to burn hydrogen instead of coal-gas is employed. J. F. B. Determination of Molecular Weights of Volatile Liquids. C. W. Porter. (J. Amer. Chem. SOC., 1912, 34, 1290-1293.)-1n the Victor Meyer, Gay-Lussac, and Hofmacn methods for determining molecular weights, a weighed quantity of the substance is volatilised a t a fixed temperature and pressure, and the volume of the vapour is determined by measurement.In other methods the volume and temperature are fixed, and the pressure is determined, or the weight of a known volume of the vapour is estimated, as in Dumas' method.The remaining possibility-namely, the estimation of the molecular weight of a substance by noting the temperature when the volume, pressure, and weight are known and fixed-has been utilised'by the author as the basis of his method. A glass bulb of about 300 C.C. capacity is immersed in a glycerol and water-bath contained in a glass jar tubulated near the bottom.A tube of small bore is sealed on each end of the bulb, and one of these passes through a stopper in the tubuiure of the jar, and is connected by rubber tubing to a mercury reservoir. The tube at the upper end of the bulb is connected byrubber tubing with a short capillary .tube, and this latter is sealed by a solid glass plug and rubber connection. The glycerol bath is electrically heated by means of coils of '' nichrome wire. The volume of the bulb from the plug at the top to a mark on the tube at the lower end is accurately determined.Mercury is poured into the reservoir until the level in the tube is at the mark just mentioned, when the bulb is completely filled with mercury by pressure applied at the top of the mercury reservoir with an air- pump.Before the mercury reaches the top of the tube, a small sealed glass bulb containing a weighed quantity of the liquid to be examined is dropped on to the mercury, so that the long capillary end of the bulb will be directed upwards. The capillary Cube fitting on to the top tube of the large glass bulb is then placed in position, held by heavy rubber tube, and wired on.Pressure is then applied till mercury appears a t the top of the capillary tube, when the solid glass plug is inserted end the rubber connection wired on. The mercury will have forced the end of the bulb holding the test-liquid into the capillary bore of the attached tube, and, by bending this slightly at the connection, the end of the small bulb may be broken off, when the vapour from the liquid will be free to escape.The bath is next heated and vigorously stirred. When the mercury has been forced down to the mark on the lower tube, the vapour in the large bulb is at atmospheric pressure, and a thermometer in the bath registers the temperature. To obtain exact measurements, the heating is continued till the level in the tube is a little lower than the mark; then the apparatus is allowed to cool slowly and the temperature read off when the level is exactly at this mark.When the rising and falling column reache8 this mark at a temperature constant to 0.5' C., the reading is recorded. The volume of the bulb must be corrected for the temperature attained, and allowance made for vapour tension by deduction from the barometric pressure. The results obtained compare favourably with those yielded by the older methods. A. R. T.596 REPORTS Modified Soxhlet Extraction Apparatus with Distillation Arrangement. F. Friedrichs. (Zeitsch. angew. Chem., 1912, 25, 2208.)-The condenser is arranged eccentrically above the extraction vessel, and can be shut off from it by rotation of the hollow ground-in stopper which closes the extraction vessel. This stopper is constricted underneath, so that it delivers the liquid in the axis of the cartridge. When extraction is complete, the stopper is rotated, and the flow of liquid from the condenser diverted, through a stopcock opened for this purpose, to 8 side flask. The extract is thus obtained freed from solvent. The whole apparatus may be fixed, as it is not necessary to detach the condenser in order to introduce or remove the substance. 0. E. M.

ISSN:0003-2654    

DOI:10.1039/AN9123700592

出版商:RSC    

年代:1912

数据来源: RSC

摘要:

596 REPORTS REPORTS. Report to the Local Government Board on the Nature of the Colouring Matter of Flour and its Relation to Processes of Natural and Artificial Bleaching. G. W. Monier-Williams. (Food Reports, No. 19, 1912.)-The present work is a continuation of that described by Hamill and the author in a previous report on the bleaching of flour (ANALYST, 1911, 36, 254). The colouring matter of flour is either carrotene (C4,,H5& or a substance, or mixture of substances, so closely allied to it that the absorption spectra are practically identical.The colour of carrotene can be discharged either by atmospheric oxygen or by nitrogen peroxide, but these two processes are distinct, and result in the formation of different sub- stances. Pure carrotene, on exposure to air, is bleached by absorption of oxygen, no oxides of nitrogen being absorbed from the air.It is only reasonable to suppose that the natural ageing of flour is a, similar process, and that in the bleaohing of flour by nitrogen peroxide certain substances are produced which are dissimilar to those formed during the natural ageing of flour. Although these substances are present in exceedingly minute amounts, it would appear that artificially bleached flour is not quite the same as flour that has been naturally aged.The amount of nitrite-reacting substance present in flour tends to increase on storage, although not to any great extent. Even in unbleached flours which had been exposed to the impure air of Widnes, highly contaminated with acid fumes, the maximum amount found was 14 parts as sodium nitrite per million.Whether the slight increase in nitrite-reacting substance is due wholly to absorption of oxides of nitrogen or nitrites from the air, or to chmges occurring in the flour itself, is not clear. I t is noteworthy that unbleached flours showed on the average an increase of nitrite from 0-4 to 1.2 parts sodium nitrite per million, while the bleached samples, which already contained 1-6 parts per million, showed no increase at allexoept in one instance.The results given by the Griess-Ilosvay test must therefore be interpreted with caution, and from the experiments recorded it appears extremely improbable that unbleached flour stored under ordinary conditions will show more than 1-5 to 2-0 parts per million of sodium nitrite by this test.The report concludes with photographic reproductions of the absorption spectra of the petroleum ether extractsREPORTS 597 of a bleached and unbleached flour, and of a solution of cnrrotene in the same solvent. Short addenda are attached-(1) on the effect of prolonged storage upon the nitrite content of highly bleached flour, and (2) on the effect of excessive bleaching upon the baking qualities of flour, H.F. E. H. Metropolitan Water Board. Sixth Annual Report. A. C. Houston.- This Report contains the results of the chemical and bacteriological examination of the London waters for the twelve months ended March 31, 1912. A considerable amount of the water abstracted from the Rivers Thames and Lea is now stored before filtration, and numerous results are given, of which the following may be cited, illustrating the effect of storage, and also the effect of the subsequent sand filtration of the stored water. Parts per 100,000.Ammoniacal nitrogen ... ... ... Albuminoid ,, ... ... ... Oxidised ... ... ... Chlorine ... ... ... ... ... 9, Oxygen absorption, three hours at 80" F. Turbiditv .... . . . . . . ... ... Colour, km. brown in 2-foot tube ... Total hardness ... ... ... ... Permanent hardness ... Bacteria per C.C. : gelatine ... ... ,' ,, agar ... 9 , ,, bile-salt agar ... ... ... ... ... Thames, before Storage. 0-0073 0-0155 0.26 1.76 0.2062 3-33 69 21 -67 5.62 9,155 356 48 Thames, after Storage. 0.0024 0.0109 0.25 1.71 0.1458 0.81 39 21.33 5.40 1,079 47 7 Thames, after Filtration.0.0002 0.0059 0.25 1.70 0.0892 Nil 19 21-00 6.09 10-9 1.7 - The author strongly emphasises the value of storage as improving the quality and increasing the safety of the water-supply. The test for the presence of spores of the B. enteritidis sprogenes is not now applied to the filtered waters, as they practically always yield negative results. In order to obtain a rapid provisional idea of the quality of the water sent into consumption each day, the five minutes' oxygen absorption test at 80" F. is made, and a result of 0.038 per 100,000 (experimentally axrived at) is taken as the standard which a sample should not exceed.The author concludes with a review of the position of the London water-supply, in which he states that the river waters are undoubtedly unsatisfactory in quality, and the judicious selection of water is becoming increasingly difficult, as the con- sumption increaees while the volume of the rivers remains constant. The levelling effect of storage, however, renders the possibility of any sudden dangerous pollution a matter of less grave concern than would otherwise be the case. But it is unwise to abstract water from the rivers independently of its quality, and therefore the adoption of certain supplementary processes of purification antecedent to filtration should be considered in the future. J. H. J.

ISSN:0003-2654    

DOI:10.1039/AN9123700596

出版商:RSC    

年代:1912

数据来源: RSC