to displace the air inside the apparatus, while в serves as an exit for the gas, dipping into a beaker containing a known volume of standardised iodine solution made alkaline with sodium bicarbonate. The contents of the flask are heated to boiling, then cooled by a current of cold water, causing condensation of steam and back-rush of the iodine solution, which must enter in sufficient quantity to retain its brownish red colour. The solutions of the flask and beaker together with the washings of the whole apparatus are collected, and the residual iodine is estimated. If an equal volume of the SnCl2 solution as used here had been previously estimated, the equivalent of iodine with respect to the weight of KCIO, taken is determined. Practical difficulties in transforming the iodine solution from the beaker to the flask may be avoided by using a distilling flask as in b, which, besides the requisites of a, carries a dropping funnel, the lower part of which contains distilled water, while the bulb is filled with a measured volume of iodised starch solution. To shorten the process further, the iodine may be replaced by SnCl2 solution itself, which is occasionally poured drop by drop into the boiling solution of NaOH and KCIO3, coloured with a few cc. of aniline hydrochloride indicator (Juan Fages Virgili's "Detection of Chlorates," as in Journ. Chem. Soc. Abs., Feb., 1909, ii., 179), until the blue colour disappears. The liquid in the dropping funnel should not be emptied, but, if necessary, a suitable quantity may be introduced from a pipette. Boiling the SnCl2 solution could hardly be expected to be an impediment, as a similar method has been adopted successfully by E. Dreyfus (as in Crookes's "Select Methods in Chemical Analysis," Fourth Edition, p. 562). Experimental verification of the above is still proceeding. will fill the whole of the lower part and half fill the tube A. The gas enters by a nozzle at C, and each bubble will push some liquid before it, so that there is a constant circulation of liquid in the direction shown by the arrows. A large amount of liquid may be used by increasing the size of the bulb B, while the resistance to the passage of gas is very small if the tube a is kept in an almost horizontal position. The third form of gas-washer was a slight modification of the Richardson wash-bottle. The Richardson pattern is probably the most efficient practical wash-bottle obtainable, but in its usual form the resistance offered to a gas is considerable. The modified apparatus is shown in Fig. 3. In the ordinary form the gas enters by a nozzle at the BA GAS-WASHING BOTTLES WITH VERY SLIGHT In the course of a research at Birkbeck College it was desired to purify hydrogen by washing with a number of different liquids. It was found that the pressure obtained with an ordinary Kipp's apparatus was barely sufficient to drive the gas through two Giessler's bulbs. Various forms of apparatus were tried, but in all cases the total pressure was too great, so that it became necessary to devise forms of washers in which the resistance was reduced to a minimum. As these may be useful to others, the three most convenient forms are described. Examination of an ordinary gas-washing bottle shows it to be most unscientific in principle. Unless a large resistance pressure is to be obtained, the amount of liquid must be kept small. In ordinary use the apparatus is filled about an inch above the inlet tube. This leaves a large air space through which the gas must be passed for a long time before the air is completely displaced and a pure supply of gas obtained. The apparatus described in Fig. 1 is a simple modification of an ordinary washingbottle in which a large quantity of liquid may be used without increase in the pressure. A short tube is joined to the inlet tube, and the end bent upwards as shown in diagram A. The apparatus must be filled above the side tube. The gas must be passed in at such a rate that each bubble as it passes up the side-tube A carries some liquid before it. There is thus a constant circulation of liquid up the tube B, from the bottom, and out through A at the top. The second form, shown in Fig. 2, is more efficient than the above, but less simple to make. The tube A should be of about 5 mm. internal diameter, and the efficiency of the washing increases with the length of this tube. The design of the apparatus is sufficiently explained by the diagram. The amount of liquid in the apparatus should be such as A Paper read before the Faraday Society, January 18, 1910. B FIG. I. bottom of the bulb, but the pressure is considerably reduced if the inlet tube A is prolonged as shown in the diagram. If at the commencement of an experiment the tube A is full of liquid, it is necessary to apply an increased pressure to clear the tube of liquid. This may usually be effected by sudden squeezing of one of the rubber connections. In order to keep the resistance low the diameters of the tubes at B must be properly adjusted. It was found that suitable sizes for these three tubes were 5, 12, and 20 mm. internal diameter. This is wider than is usual. The top of the innermost tubes should be cut off at as sharp an angle as possible to prevent the cap closing the opening. lessened by these modifications, but it still remains more The efficiency of the apparatus is doubtless somewhat efficient than other forms. The resistance is greater than in the second form described. FIG. 2. It will be noticed that in all three of these washers the dead space above the liquid is kept small, and that in each case there is a constant circulation of the liquid. THE ACTION OF ORGANIC PEROXIDES ON THE PHOTOGRAPHIC PLATE. By BENJAMIN T. BROOKS. WITH the recent development of our knowledge of radioactivity, the action of many substances on the photographic plate has been studied, and in some cases this property has been taken as a test for radio-activity. This property alone is not sufficient to characterise a body as being radioactive, as is very generally recognised. Nevertheless, papers still continue to find their way into the chemical literature ascribing radio-activity to certain substances which affect the photographic plate, but which have little or no semblance to the heavy metals. That the "pseudo-radio-active" substances are not radioactive in the same sense as the heavy metals has been shown by Russell, Saeland, and Ebler. Russell has described recently the action on the photographic plate of colophony and a number of substances which contain resin (Proc. Roy. Soc., B, 1908, lxxx., 376). He has shown that the shadows thrown by resin are not bounded by straight lines, but curve round a screen; that the action is not capable of passing through glass, mica, or aluminium foil, even of extreme thinness, and does not affect an electrical field. The action can pass along a bent glass tube, and may be swept out of a tube by a slow current of gas. No action takes place in an atmosphere of carbon dioxide. Heat destroys the activity and previous exposure to sunlight accelerates it. Alkalis or sulphur dioxide destroy the activity. When the activity of a specimen is destroyed by any of the above means, exposure to oxygen and light restores it. Russell also prepared crystalline abietic acid, and found it to be active. Turpentine and specimens of pure pinene and limonene after exposure to air showed the same behaviour. A similar behaviour of linseed-oil is described by him in a previous communication (Proc. Roy. Soc., 1898, lxiii., 102). Russell states that the effect appears to be produced by a vapour rather than by any form of radio-activity. He had previously shown that the vapours of hydrogen peroxide affect a photographic plate, even in dilutions of one part in a million (Proc. Roy. Soc., 1899, lxiv., 409). He suggests that this may be the active substance. Saeland has recently shown that the action on the photographic plate of the alkali metals, magnesium, zinc, and other metals which have a high solution tension, is Russell made no attempt to explain the formation of hydrogen peroxide by the organic substances studied by him, or to call attention to the fact that they are capable of forming organic peroxides by autoxidation. The conditions under which the activity is destroyed or increased are strongly suggestive of an intimate connection with organic peroxides. That organic peroxides are slowly hydrolysed by water with the formation of hydrogen peroxide has been shown by Freer and Novy (Am. Chem. Fourn., 1902, xxvii., 161), and by Clover and Richmond (Am. Chem. Journ., 1903, xxix., 179). This fact offers a plausible explanation of the formation of hydrogen peroxide, and the consequent change in the photographic plate produced by the substances studied by Russell. In order further to test this theory, benzoperacid, acetyl peroxide, and benzoylacetyl peroxide were tested, and each substance was found to affect a photographic plate. The experiment was carried out by placing about o'r grm. of the substance on a moist piece of filter-paper in a shallow crystallising dish, and covering the dish with a photographic plate. In the case of acetyl peroxide a distinct dark spot was obtained on developing the plate after an exposure lasting twenty minutes. Benzoylacetyl peroxide gave a distinct spot after about forty-five minutes. If the plates were exposed too long a reversal, or positive, was obtained. In a study of Manila copal it was found that the substance rapidly absorbed oxygen from the air, and that the powdered resin affected a photographic plate in the same way as colophony. Russell stated that amber was very feebly active compared with colophony. In order to ascertain if the older fossil resins could be differentiated in this way from the more recent ones, several samples of Manila and other copal resins were exposed to photographic plates. the interest recently shown in this peculiar phenomenon it is published at the present time, while the work on Manila copal will appear at a later date.-Philippine Journal of Science, iv., No. 5. STUDIES IN THE USES OF FINELY DIVIDED By S. W. PARR, T. R. ERNEST, and W. S. WILLIAMS. Southern Illinois have furnished, in very large measure, THE extensive deposits of finely divided silica found in the incentive for some work in the Department of Applied Chemistry of the University of Illinois during the last two years. This work has consisted mostly of experiments on compounds such as are formed at steam temperatures after steaming, good texture and strength. This suggested It was found that bricks made in this way possessed, a series of experiments to determine the best ratio of lime without pressure suggested itself, and with it came the to silica. Then the idea of making briquettes by moulding question as to the best conditions for maximum strength. In the first case the briquettes were made by compression cement briquette mould was used and briquettes were and tested for crushing strength, while in the second, a tested for tensile strength. In the first case the best lime while in the latter the best results were attained when 20 silica ratio was found to be 1 mol. of lime to I of silica, per cent of lime was used with 80 per cent of amorphous silica, as is shown by the following tables : between this material and lime. : About 10 grms. of each specimen of copal were pulverised plate was then placed with the film side down upon each In order to determine whether or not the resin acids in Manila copal would, like abietic acid, affect the photographic plate in the absence of terpenes, the following experiment was tried : About 50 grms. of powdered copal were dissolved in cold dilute alkali, and the solution extracted four times with ether. The solution was then evaporated to one-half its volume on the steam-bath to expel the ether. The solution was diluted, and the resin acids precipitated by dilute hydrochloric acid. Ten grms. of the dried acids were exposed to sunlight for one-half day. A photographic plate was then placed over the substance, and on developing the plate four days later a distinct black spot was obtained. The above explanation of the action on the photographic plate of certain organic substances was suggested by a study of the autoxidation of Manila copal, but in view of Exposed, 10 hrs. ΙΟ 90 134 20 80 278 30 70 204 40 60 169 Time * An examination of this material made some years ago showed it to be of chalky appearance sometimes stained by iron. It consists of minute particles of silica, from 50 mm. to o'2 mm. in diameter, of crystalline structure, transparent and irregular in shape, loosely cemented together by a small amount of clay. An analysis showed :Moisture, o 15 per cent; silica, 98.00 per cent; magnesium oxide, o‘20 per cent; aluminium oxide, 121 per cent; undetermined, 0'44 per cent. -EDITOR Journal of Industrial and Engineering Chemistry. As this behaviour is not what one would expect, the work was repeated several times with substantially the same results. After finding the best proportion in which to mix the lime and silica to give the best tensile strength, experiments were made in which other substances were introduced into the briquette. Magnesia was found to work about the same as lime. It was found that the introduction of about four parts of sharp sand into a mixture of equal parts of lime and silica (one part) gave a product with much higher tensile strength than when no sand was used. The effect of the addition of fibrous material was next studied, and the first to be tried was slag-wool. The addition of varying per cents of this material did not materially affect the strength until enough was added to dilute the original material so much that it resulted in a weakening of the bond. There appeared to be no union between the fibre and lime. A cheap grade of asbestos was tried next. The results of the tests are shown in the following table : From the table it will be observed that the tensile strength rises with the addition of asbestos until 12 per cent is added, when the drop is sudden. It will be noticed, too, that the strength of the material may be increased very materially by the addition of this fibre. The effect of the addition of colloids was next studied. To the regular mixture of lime and silica (20 per cent CaO to 80 per cent SiO2), varying per cents of sodium silicate and of Portland cement were added, and the briquettes made and tested in the usual way. The results showed a decrease in the strength in every case, which in the case of the sodium silicate was about 66 per cent of the original strength when only 5 per cent was added. Mixtures of lime and silica in the proportions for best tensile strength were found to possess good plasticity, this property being possessed in about the same degree as in the case of Georgia kaolin. This discovery led to an attempt to make a vase of this material. The regular mixture of lime and silica was used, enough water being added to give the best plasticity, and the vase moulded by hand in a plaster of Paris mould. When dry, the vase could be removed and the joints finished. When hardened by steam it had, when struck with a pencil, the true hard ring of vitrified clay biscuit. The material easily took the form and markings of the mould; there was no shrinkage that was noticeable, and the colour was a good clear white. No attempt was made to prepare a vase by making the mixture up with enough water so that it could be poured into a mould in the form of a slip, but there can be little doubt as to the possibility of making ware in this way. Small trials were also made by jiggering, which were very successful. These experiments suggest the possibility of using this process for making architectural building material to be The manufacturer of used in a way similar to terra cotta. terra cotta has many obstacles to overcome in the process of burning. His ware may warp or be of the wrong colour; it may shrink excessively or be defective in some other way. The manufacture of artificial stone that might be used similarly to terra cotta might be made, it seems, very advantageously by this process. Journal of Industrial and Engineering Chemistry, October, 1909. THE PREPARATION OF PERCHLORIC ACID FROM SODIUM PERCHLORATE. By FRANK C. MATHERS. MOST of the methods for the preparation of perchloric acid depend upon the distillation under reduced pressure of a mixture of a perchlorate and sulphuric acid, or of a crude perchloric acid that has been prepared by the decomposition of aqueous chloric acid by heat, by the action of sulphuric acid upon barium perchlorate, or by the action of hydrofluosilicic acid upon potassium perchlorate. Kreider has described a method which avoids this troublesome distillation under reduced pressure, which is easy of manipulation, and which gives a very pure perchloric acid (Am. Journ. Sci., [3], xlix., 443; Zeit. Anorg. Chem., ix., 343; Treadwell and Hall, Quantitative Chemistry," 1904, ii., 47). His method is based upon the reaction expressed by this equation:NACIO+conc. HCl(+excess conc. HCl) = NaCl + HCIƆ4. 66 Solid dry sodium perchlorate is treated with an excess of concentrated hydrochloric acid. The mixture is then filtered, and the residue of sodium chloride, which is almost insoluble in the excess of hydrochloric acid, is washed with concentrated hydrochloric acid. The filtrate is a mixture of perchloric acid, hydrochloric acid, and small amounts of the sodium salts of these acids due to the slight solubility of the sodium chloride in the hydrochloric acid solution. By heating this filtrate until white fumes of perchloric acid are evolved, the hydrochloric acid is volatilised, and the perchloric acid remains behind. The boiling points of the hydrochloric acid and the perchloric acid with 2 molecules of water (119 and 203° respectively) are so far apart that a very satisfactory separation is obtained. The object of this research was to determine the best conditions and the proper quantities of reagents to use in order to obtain the best results from this process of Kreider, since the original article only gave a qualitative description of the method. Twenty grms. of sodium perchlorate (weighed to 1 mgrm.) was placed in a 100 cc. beaker, and treated with the concentrated hydrochloric acid. The contents of the beaker were filtered upon a Gooch crucible, and the residue of sodium chloride washed with ten cc. portions of con HCI. HCIO4 in filtrate, in NaClO4 Grms. Per cent. in filtrate. Grms. Total C104 in filtrate, NaC104 in terms of NaClO4. in NaCl residue. Grms. Per cent. Grms. ΙΟ 13.65 71'4 15:56 81.4 2.72 centrated hydrochloric acid. The filtrate, which contained | TABLE II.-Effect of the Quantity of Hydrochloric Acid. the aqueous perchloric acid and the excess of hydrochloric acid together with small amounts of the sodium salts of Cc. of these acids, was evaporated upon a hot plate to volatilise the hydrochloric acid. The residue which did not volatilise below a temperature of 150° consisted of aqueous perchloric acid whose purity and yield depended upon the conditions of the experiment. These samples of perchloric acid were analysed to determine the free perchloric acid, the sodium perchlorate, and the hydrochloric acid. The residues of sodium chloride that were obtained by the first filtration upon the Gooch crucibles were analysed to determine the sodium perchlorate. Methods of Analysis. Free acids were determined by titration, using methylorange as indicator. The end-point with perchloric acid was decisive and satisfactory. Volhard's method was used for the volumetric determination of the chlorides. The perchlorates in the free perchloric acid were determined by evaporating a measured portion to dryness in a platinum dish. The dish was then heated to near redness until the perchlorates were decomposed to chlorides. The end of this decomposition was easily detected, since the perchlorates were easily fusible, and the chlorides were infusible at this temperature. The total residue, which consisted essentially of sodium chloride, was calculated to sodium perchlorate. Of course this method would only correct results with samples containing perchloric acid, sodium perchlorate, and easily volatile substances such as hydrochloric acid. The estimation of perchlorates in the presence of chlorides depends upon the determination of the difference between the chlorine in a sample that has been treated to decompose the perchlorates into chlorides, and in an untreated sample. This decomposition can be accomplished very easily by the method of Dittrich and Hollenback (Ber., 1905, xxxviii., 751). The perchlorate is fused for several hours with sodium nitrite. After cooling, the fused mass is dissolved in water, and the chlorine is determined by the method of Volhard. Porcelain dishes are attacked by the fused sodium nitrite, so platinum vessels must be used. The sodium nitrite that was used in this research contained chlorine, so a blank was determined, and the proper correction was applied to each analysis. This method gave uniformly accurate results, and was satisfactory in every way. Materials Used. A commercial preparation of sodium perchlorate was used. Its composition was :-NaCl, 1.76, 1.86; NaClO4, 95 38, 95'77 per cent; NaCIO3, trace. Commercial barium perchlorate was used. The material was "caked" in the bottle, and an average sample for analysis was difficult to obtain. An analysis showed 53.5 per cent of CIO4. The C.P. hydrochloric acid which was used showed a specific gravity (spindle) of 116 at 24°. The commercial acid, which was used in one experiment, had a specific gravity of 114 at 24°, and each cc. contained o'00097 grm. of non-volatile matter. 50 The filtrates containing the perchloric and hydrochloric acids were heated upon a hot plate. The sides of the beaker should be brought to the temperature of the experiment, since, otherwise, the drops of the liquid which condense upon the upper part of the beaker retain hydrochloric When the temperature of the entire beaker was brought to 135° there was not enough hydrochloric acid remaining to give an opalescence with silver nitrate. All of the hydrochloric acid could probably have beer. removed by maintaining the temperature somewhat below 135° for a period of time, but this experiment was not tried. Table III. shows that all of the hydrochloric acid is volatilised at a temperature of 135°. TABLE V.-Amount of Washing needed to remove the Perchloric Acid from the Sodium Chloride Residues. The residue was first drained by suction. It was then washed with five I cc. portions of concentrated hydrochloric acid, again drained by suction, then washed with a second five 1 cc. portions of acid, &c. Each five I cc. portions were saved separately and analysed. Conc. HCl used in washing. First five 1 cc. portions HCIO, washed from the residue of NaCl. Grm. Large amounts 1.65 O'19 0'08 After washing with twenty portions of acid, o'06 grm. of sodium perchlorate still remained in the residue. Table V. shows that ten I cc. portions of concentrated hydrochloric acid is the most economical amount to use in washing the residue from 20 grms. of sodium perchlorate, and 25 cc. of concentrated hydrochloric acid. |