the cloud. When no further precipitation was noticed the liquid was back-titrated with the silver nitrate solution. Finally, when complete, an excess of about 3 cc. of silver solution was added, and the flask allowed to remain for about eighteen hours in a dark room. The clear liquid was then drawn over by the capillary syphon in the manner already described (loc. cit.), and the precipitate of silver bromide washed by seven successive additions, first, of very dilute nitric acid, and then of cold water. The solution, containing mainly strontium nitrate, and the washing water were collected in a graduated cylinder, and their aggregate volume noted. Assuming that 2 grins. of strontium bromide were used, equivalent to 171 grms. strontium nitrate in the solution over the precipitated silver bromide, that this solution measured 30 cc., and that 27 cc. were drawn off at each operation, 30 cc. of washing water being uniformly added, then there would be left : After the first drawing off 0.1710 grm. Sr(NO3)2 that is after the 4th washing the amount of any admixed strontium nitrate would be inappreciable by the balance. The The washed silver bromide was then dried, first at 110°, then at 160°, and was fused by placing the flask containing it on a platinum tray heated by an alcohol flame. tare was similarly treated, and both flasks were placed side by side in a desiccator containing phosphoric oxide for fifteen to twenty hours before being weighed. Re-fusion rarely made any difference in the weight of the silver bromide; in no case did it exceed o'02 of a mgrm. The fused silver salt was always perfectly transparent, and of a faint yellow colour. The weight needs a small additive correction on account of the solubility of silver bromide. Determinations by electrical conductivity measurements of the solubility of silver bromide in water at ordinary temperatures have been made by Kohlrausch and Rose (Zeit. Phys. Chem., 1893, xii., 240), W. Böttger (Zeit. Phys. Chem., 1903, xlvi., 602), and Kohlrausch (Zeit. Phys. Chem., 1904, 1., 356). The results are not very concordant, but the mean amount may be taken, without sensible error, as o'2 mgrm. per litre. OF FOODS FOR SALICYLIC ACID. We assume that the amount is not materially affected by A SOURCE OF ERROR IN THE EXAMINATION the presence of free nitric acid or of strontium nitrate. In any case the correction is very much smaller in amount than for silver chloride, and has only a very slight influence on the value of the atomic weight. The following example is given to show the character of the data obtained, and the mode of calculation by which the resultant atomic weight was deduced :— By H. C. SHERMAN. IN testing foods for preservatives it is commonly assumed that a constituent volatile with steam, soluble in ether, capable of sublimation and crystallisation, and giving a violet reaction with ferric chloride, must be salicylic acid. This assumption is, however, incorrect. It has recently come to the attention of the writer that certain baked cereal products and prepared foods containing such products yield a constituent which shows all of these properties, and which would, therefore, upon ordinary examination, cause these foods to be reported as containing salicylic acid; but further study showed that the constituent in question did not come from the raw materials used, but was developed during the baking process. Salicylic acid was, in fact, absent, and the source of error was indicated by an examination of the literature of the ferric chloride test for salicylic acid and was fully demonstrated by experiment. Literature. Brand found, in 1893 (Zeit. Ges. Brauw., xv., 303), that an extract of caramel malt gave a reaction with ferric chloride like that given by salicylic acid, and in a paper published the following year (Ber., xxvii., 806) cites Erich go When this was mixed Source of Error in Examination of Foods for Salicylic Acid, (Der Bierbrauer, xxiv., 465) and Munsche (Woch. f. | total distillate measured 800 cc. Brauerei, x., 739) as having made the same observation. In his paper of 1894 (Ber., xxvii., 806), Brand recorded the actual separation from such caramel malt of a crystalline substance, soluble in ether or water (much less soluble in cold water than hot), volatile with steam, capable of sublimation and reacting like salicylic acid with ferric chloride, but which did not give the red reaction of salicylic acid with Millon's reagent. By condensing the vapours given off during the roasting process in the manufacture of a so-called malt coffee, extracting with ether, purifying and re-crystallising in different ways, Brand obtained the reacting substance in crystalline form both as leaflets and as needles. Brand named the substance "maltol," and showed that it had phenolic character and agreed with the formula C6H603. Kiliani and Bazlen (Ber., xxvii., 3115) also prepared maltol in crystalline form, showed that it has an acid reaction, and studied its chemical behaviour. Will studied its effect upon yeast (Zeit. Ges. Brauw., xxi., 307; Abs. Zeit. Unters. Nahr. Genussm., i., 720). Abraham (Journ. de Pharmacie, de Liege, 1898, v., 173; Abstracted in Repert Pharm., [3], X., 347; and Zett. Unters. Nahr. Genussm., i., 857) pointed out that dark coloured beers may readily contain maltol, which, in the examination of the beer for preservatives, will behave like salicylic acid, both in steam distillation and extraction with ether, and give a marked violet coloration with ferric chloride. The ferric chloride reaction being thus inapplicable, and the Millon reaction not suitable, Abraham recommended the Jorissen reaction. This reaction for salicylic acid, proposed by Jorissen in 1882 (Bulletins de l'Academie Royal des Sciences, des Lettres, et des Beaux-arts de Belgique, 3rd series, iii., 259), is as follows:- Jorissen's Reactions.-To the solution to be tested add four or five drops of a 10 per cent solution of potassium (or sodium) nitrite, four or five drops of acetic acid, one drop of a 10 per cent solution of copper sulphate, and heat to boiling. In the presence of salicylic acid the solution turns reddish, and with more than a very minute amount becomes blood-red. According to Jorissen phenol behaves in the same way, but benzoic acid does not. Abraham found that maltol does not give this reaction, and recommended it as the most reliable test for salicylic acid. Da Silva used Jorissen's reaction among others (Comptes Rendus, 1900, cxxxi., 423; Bull. Soc. Chim., 1900, [3], xxiii., 795). Klett (Pharm. Centr., xli., 452; Zeit. Unters. Nahr. Genussm., iv., 469) recommends Jorissen's reaction for the detection of salicylic acid in substances containing citric acid. Jorissen's reaction was one of those used by Portes and Desmoulieres (Annal. Chim. Analyt., 1901, vi., 401; Zeit. Unters. Nahr. Genussm., v., 468) as demonstrating the presence of small amounts of salicylic acid in strawberries. Windsch (Zeit. Unters. Nahr. Genussm., vi., 447) investigating the natural occurrence of salicylic acid in fruits, and having obtained violet reactions with ferric chloride from raspberries and strawberries, adopted the Jorissen reaction to decide the question whether the reacting substance was actually salicylic acid. Thus it is evident that the Jorissen reaction for the detection of salicylic acid is fairly well established in Europe. Experimental. Fifty grms. of a prepared food consisting in part of a baked cereal product were mixed with water, acidulated with phosphoric acid, and distilled with steam in an Erlenmeyer flask surrounded by oil heated to 120° to 130°. Portions of distillate tested from time to time gave distinct violet reaction with ferric chloride, but no reddish colour in Jorissen's test. The distillation was continued until the ferric chloride reaction was no longer obtained, when the and one hundredth of it (8 cc.) tested with ferric chloride a distinct reaction like that of salicylic acid was obtained. That the reacting substance was not salicylic acid was shown by extracting thoroughly the other ninety-ninehundredths of the distillate with ether, treating it as recommended by the Association of Official Agricultural Chemists (Bull. 107 (Revised), Bur. Chem., U.S. Dept. Agriculture) for the complete recovery of salicylic acid, subliming, and testing the product obtained. Although this represented practically one hundred times as much of the substance as had last been tested and found to react with ferric chloride, no reaction whatever was obtained with the Jorissen test. Comparative tests with known solutions indicated that in the writer's hands about twice as much salicylic acid is required for the Jorissen reaction as for the reaction with ferric chloride, whereas in the above case a distinct coloration was obtained with ferric chloride in a test representing o'5 grm. of the sample, and no coloration when the Jorissen reaction was applied to a solution of about equal volume, which represented about 50 grms. of the same sample, or about one hundred times as much. Hence if the reaction with ferric chloride in the first case had been due to salicylic acid there would have been present, in the second case, about fifty times as much as is required to give a distinct coloration with the Jorissen reaction. It is, therefore, abundantly evident that the substance which gave the violet reaction with ferric chloride was not salicylic acid. It was probably Brand's maltol, or some very similar substance. The production of this substance is evidently due to the baking process or caramelisation rather than to the presence of malt, though it may be produced more abundantly in the baking of malt than in the baking of flour. When 25 grms. of the brown outer crust of ordinary wheat bread (partly from "patent" and partly from "whole wheat" flour) were treated in the same manner as the prepared food described above, the distillate gave a distinct violet reaction with ferric chloride, though it was, of course, free from salicylic acid and did not give the Jorissen reaction. It may also be noted that Abraham (loc. cit.) obtained the violet (maltol) reaction with ferric chloride from roasted coffee as well as from caramel malt. Summary. Maltol, an acid substance volatile with steam, soluble in ether, capable of sublimation, crystallising sometimes in leaflets and sometimes in needles, and giving with ferric chloride the same violet colour as salicylic acid, has been shown to be formed in the baking or roasting of malt, and the same or some similar substance is produced in the roasting of coffee and the baking of wheat bread. Such a substance would be reported as salicylic acid when the present ferric chloride test is used, no matter how carefully the test be made, and even if the precaution be taken to sublime the acid and a crystalline sublimate be obtained. In order to avoid this source of error some reaction other than that with ferric chloride must be used for the detection of salicylic acid in foods. Jorissen's reaction, already well established in European literature, is not subject to the same source of error. In carrying out this reaction about 10 cc. of the liquid to be tested were used, and to this were added four or five drops of 10 per cent potassium nitrite, four or five drops of 50 per cent acetic acid, one drop of 10 per cent copper sulphate; the liquid, after shaking, was then heated to boiling, and, if necessary, boiled for half a minute, and allowed to stand one or two minutes for the red colour to develop. Under these conditions 0.00005 grm. of salicylic acid in 10 cc. water (1: 200,000) gives a distinct reddish colour, easily recog nisable after a very little practice, especially on comparison with a blank test carefully made with the same amounts as NEWS the reagents. With larger amounts of salicylic acid an unmistakable blood-red colour develops quickly on heating. -Journal of Industrial and Engineering Chemistry, ii., No. I. provement both in rapidity and accuracy over the use of THE DETERMINATION OF IRON AND ALUMINA THAT the estimation of iron and alumina in the presence of phosphoric acid, manganese, calcium, and magnesium is a somewhat troublesome operation is evidenced by the fact that the American Association of Official Agricultural Chemists has adopted no method for the determination of these two elements in inorganic plant constituents. The method which has been in use for some time in the laboratory of the New Mexico Agricultural College has been found quite satisfactory; and, as it contains some modifications that we have not seen suggested as applicable to the analysis of ash, we have briefly outlined it below. Separation of the Iron and Alumina.—To an aliquot of a solution of the ash corresponding to 1 grm. add pure ferric chloride of known strength more than sufficient to combine with all phosphoric acid present. To the cold solution add concentrated solution of sodium carbonate until permanent precipitate just begins to form. Clear with 1 cc. of 80 per cent acetic acid, add 1 grm. of sodium acetate, and boil for three or four minutes. When the precipitate has settled sufficiently to determine that the solution is colour less, filter at once before the precipitate becomes slimy. To ensure the removal of all the lime from this precipitate it should be dissolved with hydrochloric acid into the same beaker, thereby preventing the necessity of washing beaker and precipitate. The iron and alumina are then re-precipitated with ammonia, filtered, dried, ignited, and weighed. The weight obtained, minus the F2O3 and P2Os found by other methods, and the known amount of Fe2O3 added, will be the weight of the Al2O3 in the ash. In this method a great deal depends upon the proper adjustment of the sodium carbonate and the acetic acid. If too much sodium carbonate is used, lime and magnesia may precipitate. Too much acetic acid may prevent the proper precipitation of iron and alumina. The solution of ferric chloride which is added produces a red precipitate that is easier to filter and wash, makes a clearer filtrate, and ensures a complete, precipitation of the phosphoric acid. THE DETERMINATION OF PHOSPHORUS IN By P. F. TROWBRIDGE. DURING the winter of 1907-8 the author attempted to A large platinum dish was used, and the ignition of the The method which we had adopted for the phosphorus The phosphorus was determined on the ash by the usual For the seventy-two samples the method by ashing gives In most cases the whole six determinations agree so Determination of Iron.-To an aliquot of a solution. of The Chem. Soc., 1908, xxx., 617), too much care cannot be exercised in the digestion of the ash so as to put all the phosphorus in a form that will be precipitated by the ammonium molybdate. This seems to be most surely accomplished by the Neumann (Fourn. Am. Chem. Soc., 1902, xxiv., 1106) method of digestion with equal volume of sulphuric and nitric acids; or by long (eight to ten hours) digestion with nitrohydrochloric acid. Credit is here given Mr. Norman Hendrickson for his assistance in the laboratory work.-Journal of Industrial and Engineering Chemistry, i., No. 9. PROCEEDINGS OF SOCIETIES. ROYAL SOCIETY. Ordinary Meeting, February 10th, 1910. SIR ARCHIBALD GEIKIE, K.C.B., President, in the Chair. PAPERS were read as follows: "Some Phenomena of Magnetic Disturbances at Kew." By Dr. C. CHREE, F.R.S. A recent paper (Phil. Trans., A, ccviii., 205) discussed the diurnal inequality of Kew magnetic declination derived from 209 of the most highly disturbed days of the eleven years 1890 to 1900. The present paper discusses the corresponding phenomena for the same days in the other magnetic elements. It is shown that the irregular changes which form the most obvious feature of magnetic storms are accompanied by large regular diurnal changes, which are specially striking in the vertical force. In this element the disturbed days referred to above gave a regular diurnal inequality, whose range in the average month of the year was about four times that given by the Astronomer Royal's "quiet" days. The influence of the hour of the day on the character of the disturbance is visible even on casual inspection of the vertical force curves. When disturbances lasting only a few hours occur in the late afternoon, there is almost invariably a rise in the force, whereas when they occur in the early morning there is a fall. Besides dealing with the analysis of the diurnal inequalities derived from the disturbed day curves, the paper discusses some new phenomena observed in the a-periodic changes of the magnetic elements. "Novel Phenomenon in the Diurnal Inequality of Terrestrial Magnetism at Certain Stations." By R. B. SANGSTER. NEWS Average x. of g, h, i. 02262 Average of Total v, w, x. average. 02232 0*2247 02081 0.2158 0*2120 01927 01943 01935 01897 0 1892 0*1945 The author has measured the wave-lengths of these lines, so that now 24 potassium lines, 25 rubidium lines, and 19 cæsium lines are known of the principal series. Of these, 15 are new in the case of potassium, 21 in the case of rubidium, and 12 in the case of cæsium.. In the cases of rubidium and cæsium, the metals themselves were not available, but by heating the chlorides with sodium or potassium, enough vapour was obtained to show the absorption spectrum quite definitely. These lines, with the lines measured by Wood for sodium, give good data for testing various formula that have been suggested for representing the series lines. None of the suggested formulæ tested give values representing the series within the limits of experimental error. In particular, the quantity of Rydberg's formula No, or of the modified Rydberg formula of Ritz, is shown not to be constant. One of the most interesting facts arising out of the investigation is that none of the lines of the associated series appear in these absorption spectra. Channelled space spectra appear which are analogous to the similar spectra for sodium vapour. Further interesting facts noted are in regard to the effect of mixtures of vapours. Some lines or bands appear in spectra of mixtures which are apparently unconnected with the spectra of either constituent. This was specially evident in the case of cæsium and sodium; a set of bands appeared at about W.L. 3000-3500 which do not appear in the sodium spectrum, nor in the mixture of potassium and cæsium spectrum. Other interesting phenomena appear as the density of the vapour is increased in the widening of the lines and the appearance of satellites connected with the lines of the series. vapour of lithium has not yet been successfully investigated, as it attacks the material of all tubes hitherto tried. The "Shapes of the Isogeotherms under Mountain Ranges in Radio-active Districts." By Prof. C. H. LEES, F.R.S. By F. B. PIDDUCK, B.A. "Propagation of a Disturbance in a Fluid under Gravity." "Flow of Water through Pipes and Passages having Converging or Diverging Boundaries." By Dr. A. H. GIBSON. "Effect of Pressure upon Arc Spectra: Titanium." By R. Rossi. The work is on the range from λ4000 to λ4600, examined with the 21 feet concave grating spectrograph of the Manchester University Physical Laboratory, and which gives on the photographic plate a dispersion of 1'3 Angström units per millimetre. The arc was formed between a carbon pole and a graphite tube filled with titanium carbide. The pressures at which the photographs were taken were 15, 30, 50, and 100 atmospheres. The broadening, reversal, displacement, and changes of relative intensity of 52 lines were studied. All lines were found to broaden out with an increase of pressure, the amount and type of broadening being different for different lines. Several lines were found to reverse under pressure, some symmetrically and some asymmetrically. All lines were found to be displaced towards the red end of the spectrum, the displacement being a linear function of the pressure within the limits of accuracy of experiment. The value of the displacement varies for different lines, and the unreversed lines cannot be grouped into sets giving the same displacement. The reversed lines, however, with the exception of one, can be formed into two groups, their mean displacements being very nearly in the ratio 3:5. The mean displacement per atmosphere of all the titanium lines studied is found to be 0.003652 Angström units. The limited number of lines studied, both in this work and by other workers on the Zeeman effect, do not enable one to obtain any relation between the pressure displacement and magnetic separation. The relative intensity in nearly all lines is altered by pressure, and a list is given of the lines which are thus enhanced or weakened. "Change of Carbon Disulphide into a Gaseous Product Condensable and Explosive near the Temperature of Liquid Air" (see p. 86). By Sir JAMES Dewar, F.R.S., and Dr. H. O. JONES. PHYSICAL SOCIETY. Annual General Meeting, February 11th, 1910. Dr. C. CHREE, F.R.S., President, in the Chair. THE report of the Council was read by the Secretary. Report of the Council. Since the last Annual General Meeting, eleven ordinary Science Meetings and one informal Meeting of the Society have been held at the Imperial College of Science and Technology. The average attendance at the Meetings, apart from the Annual Exhibition, has been forty-one as compared with thirty-seven last session. The Fifth Annual Exhibition of Apparatus by Manufacturers was held on December 14th. A departure from the usual programme was made on this occasion. The Exhibition was held on a Tuesday instead of a Friday, so that it might be open on both the afternoon and evening, and experimental demonstrations were given by Prof. C. V. Boys, F.R.S., and Prof. S. P. Thompson, F.R.S. There were thirty-seven exhibitors. The new arrangements seem to have been appreciated, as there was a very much larger attendance than hitherto, the number of Fellows and visitors (including exhibitors) amounting to about 700. The number of Ordinary Fellows now on the roll, as distinct from Honorary Fellows, is 436, an increase of 6 on the number last year; 20 new Fellows have been elected. Two Honorary Fellows were elected at the last Annual General Meeting, namely, Mons. R. Benoit and Prof. Julius Thomsen. There have been four resignations, and five Fellows have been struck off the register for nonpayment of subscriptions. The Society has to mourn the loss by death of two Honorary Fellows, namely, Prof. F. Kohlrausch and Prof. Simon Newcomb; one Past Presi dent, Dr. Shelford Bidwell, and four other Ordinary Fellows, namely, H. E. Harrison, Ludwig Mond, F.R.S., W. J. Russell, F.R.S., and W. F. Stanley, to the last of whom the Society in indebted for funds enabling them to issue the Bulletin. The Treasurer's report was read by the TREASURER as follows: Treasurer's Report. Possibly owing to the amendment of Article 26 of the Memorandum of Association, the year 1909 has been a record year for the receipt of subscriptions, the amount collected exceeding that realised in 1908 by over £150. At the same time the printer's bill has swollen in a similar proportion, owing to the Secretary's efforts to expedite publication, so that the net balance on current account has increased by only £22. The assets of the Society show The following Officers and Council were elected for the ensuing year — President-Prof. H. L. Callendar, M.A., F.R.S. Vice-Presidents-Those who have filled the Office of President, together with A. Campbell, B.A.; Prof. C. H. Lees, D.Sc., F.R.S.; Prof. A. Schuster, Ph.D., F.R.S. ; S. Skinner, M.A. Secretaries-W. R. Cooper, M.A., and S. W. J. Smith, M.A., D.Sc. Foreign Secretary-Prof. S. P. Thompson, D.Sc., F.R.S. Treasurer-W. Duddell, F.R.S. Librarian-W. Watson, D.Sc., F.R.S. Other Members of Council-Prof. W. H. Bragg, M.A., F.R.S.; W. H. Eccles, D.Sc.; A. Griffiths, D.Sc.; J. A. Harker, D.Sc.; Prof. T. Mather, F.R.S.; A. Russell, M.A., D.Sc.; W. N. Shaw, M.A., F.R.S.; F. E. Smith ; R. S. Whipple; and R. S. Willows, M.A., D.Sc. Prof. H. L. CALLENDAR then took the Chair and delivered an Address. current. After referring briefly to the losses sustained by the Society since the last General Meeting, the President dealt with the application of resistance thermometers to the recording of clinical temperatures. Records of clinical temperatures have been obtained by Dr. Gamgee using a thermocouple in conjunction with sensitive recording inIn the first place, the E.M.F.'s developed are so small struments. The objections to thermocouples are twofold. that the recording instruments must be very sensitive, and therefore unsuitable for ordinary use. serious difficulties arise with regard to the thermostat In the second place, which is necessary to maintain one of the junctions of the thermocouple at a constant temperature. The chief difficulty in connection with the use of resistance thermometers for this kind of work lies in the heating effect of the In order to explain how this difficulty is overcome Prof. Callendar gave a short account of the conditions upon which the sensitiveness of a Wheatstone's bridge depends. He pointed out that in platinum thermometry, of the leads, it is necessary that the ratio arms of the in order to obtain accurate compensation for the resistance bridge should be equal, and he showed that this condition reduced the sensitiveness, which could be obtained by suitably varying the resistances by about 30 per cent. In joining up a bridge in work with resistance thermometers, Maxwell's rule for the positions of the battery and galvanometer which give maximum sensitiveness is seldom applicable. While Maxwell's arrangement actually gives the greatest sensitiveness, the heating effect of the current is so much greater that the trouble arising from this cause more than counterbalances the increased sensitiveness. The problem to be solved in designing a suitable thermometer for clinical work is, with a given galvanometer and which will give the most accurate results for a given resistance-box, to find the resistance of the thermometer heating effect of the current. This is given by the equation R=2G+S, where G is the resistance of the galvanometer and S that of one of the ratio arms. In the apparatus used at the meeting G=S=10 ohms giving R=30 ohms, a resistance for the thermometer which gives the convenient scale of o.1 ohm increase per degree Centigrade. It is important in the construction of a thermometer for clinical work to secure quickness of action and to reduce the heating effect of the current. An ordinary tube-form of thermometer is good for laboratory work with sensitive galvanometers, but it is unsuitable for use with recorders. An ordinary form of thermometer takes about one and a-half minutes to reach a steady temperature when taken from water at the temperature of the room and placed in water at o° C., whereas a thermometer suitably designed will |