CHEMICAL NEWS, Feb. 17, 1911 New Derivatives of d-Glucosamine. The results obtained, however, did not permit of any generalisation being put forward, owing to the fact that anomalies were observed in some cases which at present have not been satisfactorily explained. The following compounds were described:--Nitro-o-, m-, and p-hydroxystyrene, OH C6H4 CH: CH NO2 (m. p. 133-134°, 132-133°, and 165° respectively); w-nitro-pacetoxystyrene (m. p. 158—159°); w-nitro-p-benzoyloxystyrene (m. p. 156°); ethyl w-nitrostyryl-p-oxy acetate, NO2 CH CH C6H4 O·CH2 CO2Et (m. p. 82°); ethyl o nitrostyryl-m-oxyacetate (m. p. 147-148°); p-chloro- ·0)nitrostyrene (m. p. 111-112); w-3-dinitro-2-hydroxystyrene (m. p. 174-175°); w-3-dinitro-4-methoxystyrene (m. p. 165-166°); 3-3 5-trinitro-4-methoxy-a-phenylethanol, OMe C6H2(NO2)2°CH(OH) CH2 NO2; B-nitro2:5-dihydroxyphenylethanol (m. p. 182°); ethyl 3-aldehydosalicylate, CHO C6H3(OH) CO2Et (m. p. 66—67°); ethyl 5-aldehydosalicylate (m. p. 69°), and its 3-nitroderivative (m. p. 79-80°); w-nitro-2-hydroxy-3-carboxystyrene (m. p. 231); 3-nitro-5-aldehy dosalicylic acid (m. p. 177°); 5-nitro-3-aldehydosalicylic acid (m. p. 195– 196°), and 1-w-nitrovinyl-3-naphthol (m. p. 187—1889). 32. "The Phosphoric Acids." By ALFRED HOLT and JAMES ECKERSLEY MYERS. The experiments carried out by the authors lead to the following conclusions:1. Pyrophosphoric acid is formed as an intermediate compound during the hydration of metaphosphoric acid. 2. The rate of hydration does not accord with any simple order of reaction. 3. Meta- and pyro-phosphoric acids, when prepared by dehydrating the ortho-acid, give complex molecules in solution, but when prepared by decomposing the corresponding lead salts, simpler molecules result. 79 Methyammonium nitrate has a pale greenish yellow colour, and is extremely deliquescent. It decomposes when kept in a desiccator, and when heated, methyl alcohol and nitrogen are produced together with some secondary products of decomposition. 36. "The Picraminobenzoic Acids and their Salts." By JAMES CODRINGTON CROCKER and FRANK MATTHEWS. The three picraminobenzoic acids and the salts derived from them were described. The ordinary salts are generally of the "acid" type, containing two equivalents of the acid radicle to one of the base. The "semi " or acid salts have the general formula CO2M C6H4 NHR, CO2H C6H4 NHR (where R denotes the picramyl group). This type of salt is only moderately soluble in water, and in most cases is completely hydrolysed even in the cold by excess of water, with precipitation of the acid, alkali remaining in solution. The "normal" salts are easily soluble in water, and show little hydrolysis in solution. The following substances were prepared :— o-Picraminobenzoic acid (m. p. 270°) and the acid o-picraminobenzoates of sodium, potassium, ammonium, methylamine, ethylamine, aniline, and pyridine; also the normal silver salt of o-picraminobenzoic acid. m-Picraminobenzoic acid (m. p. 234°) and the acid m picraminobenzoates of sodium, potassium, methylamine (two salts), ethylamine, and guanidine; also the normal potassium salt of m-picraminobenzoic acid. p-Picraminobenzoic acid (m. p. 285°) and the acid p-picraminobenzoates of potassium and methylamine, and the normal salts of sodium and ammonium. Quantitative experiments on the hydrolysis by water of the acid potassium salt of the o-acid were also carried out. It was shown that when the salt was in excess, equilibrium was attained when a certain constant con33. "The Determination of Solubility Coefficients by centration of alkali in solution was reached (about N/250 Aspiration." By WILLIAM JACOB Jones. 34. "a-Amino-a-phenylacetamide and some of its Derivatives." By CHARLES HUGH CLARKE and FRANCIS FRANCIS. a-Amino-a-phenylacetamide, C6H5 CH(NH2) CO NH2, is easily obtained, in the form of its benzylidene derivative, by the action of potassium hydroxide solution and ammonia on benzaldehydecyanohydrin in the presence of benzaldehyde. From this condensation product the free base itself is isolated by the action of phenylhydrazine with subsequent separation, based on the solubility of a aminoa-phenylacetamide in water. a-Amino-a-phenylacetamide is a strong base, and easily condenses with aldehydes. Its carbethoxy-derivative lends itself so readily to the formation of hydantoin that this constitutes the best method for the formation of substances of this class. On the other hand, no derivatives of the base could be prepared from which it was possible to isolate sixmembered cyclic substances. at 25°). 37. "New Derivatives of d-Glucosamine." BY JAMES COLQUHOUN IRVINE, DAVID MCNICOLL, and ALEXANDER HYND. Glucosamine hydrochloride reacts with acetyl bromide to give bromotriacetylglucosamine hydrobromide (m. p. 149— 150°, ["]D +148'4°), which serves as the starting-point in the preparation of a number of glucosamine derivatives. The compound exists in a- and 3-forms, shows normal mutarotation when dissolved in anhydrous acetone, and condenses readily with hydroxy-compounds. When acted on by traces of water, it is completely converted into glucosamine hydrobromide, owing to hydrolysis and the slow removal of the acetyl groups. By condensation with methyl alcohol, in the presence of pyridine, it gives triacetylmethylglucosamine hydrobromide (m. p. 230-233°), which has [4]D + 20-2° in methyl-alcoholic solution, and behaves as a 3-glucoside. The acetyl groups may be eliminated from this compound by boiling with barium hydroxide, or by the action of dilute alcoholic hydrogen bromide in the cold, the reaction giving methylglucosamine as a strongly alkaline glucosidic syrup. The hydrochloride of the latter appears to be a mixture of a- and 3-stereoisomeric forms (m. p. 185-187, [a] in water 19.5°); the compound closely resembles the amino methylglucoside hydrochloride recently obtained by Fischer (Ber., 1911, xliv., 132), and is possibly identical with it. The substance is, however, remarkably stable towards hydrolytic agents, but is converted into glucosamine hydrochloride by the action of concentrated hydrochloric acid. By similar treatment, all the compounds described give glucosamine salts, and are thus true derivatives of glucosamine. The conclusion is also drawn that the use of pyridine and silver salts in the preparation of the compounds has no effect on the configuration or on the amino-group. The work is being continued with the object of correlating glucosamine with glucose, and of preparing sugar complexes containing the glucosamine residue. 38. "The Interaction of Silver Nitrate and Potassium | diminishes the speed of rotation owing to the increase of L Persulphate and its Catalytic Effect in the Oxidation of Organic Substances." By PERCY CORLETT AUSTIN. A detailed analysis of the black precipitate, which results from the interaction of aqueous solutions of silver nitrate and potassium persulphate, was found to verify the assumption made by Marshall (Proc. Roy. Soc. Edin., 1900, xxiii., 163) that the silver persulphate, first formed, is partly decomposed by water according to the equation Ag2S2O8 + 2H2O=Ag2O2 + 2H2SO4. The precipitate appears to be a mixture of silver dioxide, silver persulphate, and a small percentage of water in proportions which vary slightly according to the conditions of the experiment. The catalytic effect of adding a trace of silver nitrate in> or < may be obtained according to the resistance. the oxidation of various organic substances by means of aqueous potassium persulphate was studied. Whereas, in the absence of silver nitrate, oxidation did not appear to take place, the presence of a trace of this catalyst produced a very marked effect with toluene, when benzaldehyde and benzoic acid were quickly formed. Under similar conditions, thymol gave dithymol. and also to the fact that the hysteresis in the iron causes the secondary circuit to lag still more behind the primary. Lags greater than may be obtained when the rotation is in the opposite direction, and may be reduced to nothing by increasing the resistance. A steel core will produce a greater negative rotation than an iron one. To demonstrate the hysteresis effect it is necessary that the core should consist of a bundle of fine wires, otherwise the Foucault currents set up will introduce a lag. The effect of Foucault currents can be shown by introducing another coil within the transformer in place of the iron core and closing its circuit with a variable resistance when lags 39. "The Isomerism of Ferrocyanides." By SAMUEL HENRY CLIFFORD BRIGGS. The author finds that ferrocyanides exist in two forms, which are related to each other in the same way as are the a- and 3-ferricyanides of Locke and Edwards (Am. Chem. Journ., 1899, xxi., 193, 413), and which have therefore been termed a- and B-ferrocyanides. The a-ferrocyanides of the alkali metals are lemon-yellow in colour, and the B-ferrocyanides are orange- or amber-coloured. The difference in colour persists in solution, and when one form is converted into the other in solution, a corresponding change in colour is observed. Acids convert the a-ferrocyanides into 3-ferrocyanides; cyanides, ammonia, and alkalis effect the opposite transformation. Nine ferrocyanides have been obtained in the two modifications. The a- and 3-ferrocyanides of 1-methylammonium have [a] D 17 - 42° and - 28° respectively in absolute alcoholic solution. The conversion of the a-ferrocyanide into the 3-ferrocyanide is accompanied by a corresponding fall in specific rotation. The a- and 3-ferrocyanides are related to each other in the same way as the green and yellow platinocyanides of Levy (Trans., 1908, xciii., 1446). Reasons were adduced to show that the a- and B-ferrocyanides are stereoisomeric in accordance with the author's theory of complex salts (Trans., 1908, xciii., 1564). PHYSICAL SOCIETY. Ordinary Meeting, January 27th, 1911. Prof. C. H. LEES, F.R.S., Vice-President, in the Chair. "A Demonstration of the Phase Difference between the Primary and the Secondary Currents of a Transformer by means of Simple Apparatus," was given by Prof. F. T. TROUTON, F.R.S. The apparatus is a primitive induction motor consisting of two horseshoe electromagnets with their axes coincident and vertical, and their planes at right angles. Above the poles a copper disc is pivoted. The primary current from a transformer is sent through one magnet and the secondary current through the other. With a suitable phase difference a rotating magnetic field is thus obtained. The phase difference between the currents is The hysteresis is greatly modified by subjecting the core to a permanent magnetic field by means of an additiona coil on the transformer. Moderate fields increase the hysteresis, though large fields decrease it. This is due to working on the knee of the hysteresis curve in the first case, while in the second case the core is permanently saturated, and acts as a non-magnetic substance. Dr. W. E. SUMPNER pointed out that introducing an iron core produced several effects; for instance, the magnitudes of the currents were altered as well as the sel and mutual inductions. Prof. S. P. THOMPSON said that Mr. Baily showed a similar apparatus at the Physical Society thirty years ago. Prof. G. W. O. Howe pointed out that Foucault currents would be induced in the closed coil used for magnetising the iron core. Mr. A. CAMPBELL called attention to an instrument by means of which Lissajou figures were obtained from the combined oscillations of the primary and secondary currents. The AUTHOR, in reply to Prof. Howe, stated that a coil with a high resistance was employed for magnetising the core, to avoid the effect he mentioned. A Paper entitled, "A Note on the Experimental Measurement of the High-frequency Resistance of Wires," was read by Prof. J. A. FLEMING, F.R.S. The author refers to a paper read by him in December, 1909, before the Institution of Electrical Engineers on "Quantitative Measurements in Connection with Radiotelegraphy" (Fourn. Inst. Elect. Eng., 1910, xliv., 349), in which he described an apparatus consisting of a dif ferential air thermometer having tubular bulbs into which similar wires could be placed, and by means of which a comparison could be made of the high frequency (H.F.) resistance R' of a straight wire and its steady or ohmic resistance R. If two equal wires have passed through one, a steady current A and through the other a H.F. current A1, then if these currents are adjusted until the rate of heat evolution in each case is the same, we have A'R=A,aR'. Certain precautions are described in the paper for eliminating inequalities, but by means of correct reading H.F. ammeters as devised by the author, the ratio of the reresistances R'/R can be determined from the ratio of the mean square currents A2/A1'. The H.F. currents used were obtained by condenser discharges and the equiheating steady current determined by means of the differential thermometer arrangement having two equal wires in the two tubular bulbs. It is then shown that the results for straight wires agree very well with the formulæ given by Lord Kelvin and by Dr. A. Russell for the ratio R'/R for wires of different sizes and for different frequency. It is pointed out that the correction to be applied for damped oscillations, as compared with persistent oscillations is at most I per cent in the cases measured, and that the correction for the heating effect of the condenser charging current is negligible. The case o spiral wires is then discussed. The resistance R" of a spiral is greater than that of the same wire R' stretched out straight. In the cases examined the ratio R"/R exceeds R'/R by about 50 to 80 per cent. A formula given by Dr. Nicholson for R/R is then discussed, and CHEMICAL NEWS, Feb. 17, 1911 Chemistry of the Lead Chamber Process. is shown that it does not agree with the results of observation. Experiments are also described on the H.F. resistance of wires of magnetic metals, and it is shown that in this case the observed value of R'/R can be used to determine the permeability for small H.F. magnetising forces. Dr. A. RUSSELL pointed out that in obtaining the formula for the high-frequency resistance of iron wires it is not justifiable to assume μ to be constant. Mr. DUDDELL remarked that he had made some experiments with insulated wires, and had found that this increased the resistance, possibly owing to condenser action. Dr. ERSKINE-MURRAY said the instrument would certainly be useful to those interested in wireless telegraphy. Major O'MEARA thought the method was well worth following up. Prof. E. WILSON drew attention to the inductive action between the two circuits, which might not be negligible in the case of a spiral. A paper on the "Measurements of Energy Losses in Condensers Traversed by High Frequency Electric Oscillations," by Prof. J. A. FLEMING and Mr. G. B. DYKE, was read by Prof. FLEMING. In this paper an arrangement of apparatus is described for the purpose of measuring the internal energy losses in condensers traversed by high-frequency (H.F.) currents. It is shown. that these energy losses in condensers may be considered as if they were due to a resistance loss in a hypothetical resistance in series with the condenser, the condenser itself being supposed to have a perfect nondissipative dielectric of the same dielectric constant. This hypothetical resistance is not constant, but is a function of she condenser current. The experiments were conducted by the use of a special form of impact dischargers comprising two flat plates immersed in oil, one stationary and the other revolving at a high speed. This discharger was placed in series with a primary circuit and condenser, and H.F. oscillations were set up in the primary having any desired frequency. A secondary circuit loosely coupled consisted of a wire whose H.F. resistance could be determined, the condenser to be examined and a hot-wire ammeter and variable resistances. The measurements consisted in observing the reading of the ammeter, and then changing the current created in the secondary circuit by a small amount by adding a known resistance which altered the decrement of the circuit, but not its inductance. From these readings an equation is obtained for the hypothetical condenser resistance. It is shown that the product of the square of the secondary current A and the total resistance R of the secondary circuit is constant, and hence that the unknown condenser resistance p can be found from observations of the change in A when an additional resistance r is interpolated in the condenser circuit. The energy loss in the condenser is then Ap watts. Condensers with various dielectrics, air, oil, glass, and ebonite, were examined, and the dielectric energy losses D are stated in micro watts per cubic centimetre of dielectric for given values of the electric force E. It is shown thus D can be expressed as a function of E in the form D=XEY, where X is a constant depending on the current density, and Y is a constant depending on the nature of the dielectric. For oil and air these power losses are relatively small, but for glass and ebonite large. The necessity for measuring these losses in the case of radio-telegraphic condensers is emphasised. 2 A paper on "Some Resonance Curves taken with pact and Spark Ball Dischargers," by Prof. J. A. FLEMING and Mr. G. B. DYKE, was read by Prof. FLEMING. 81 in each case, and this has accordingly been done with both the impact and spark ball dischargers in the primary circuit and for various resistances in the secondary circuit. The results are interesting as showing exactly what takes place in each case in the primary circuit. If we are using the spark ball discharger, and if the primary and secondary circuits are coupled with various degrees of coupling, then, for any close coupling, we find on the resonance curve three peaks which correspond respectively, as regards frequency, with the free oscillation period of the secondary and with the two-period oscillations set up by the reaction of the secondary upon the primary circuit. As the coupling is weakened the double-period oscillations die out, and only the free oscillation of the secondary survives. There is always a certain coupling, not far from 10 per cent, which gives the maximum current in the secondary circuit in the form of a free oscillation. If the secondary circuit is more highly damped, then the two period oscillations are more strongly marked, and the maximum free period oscillation has a lesser maximum value. oscillations are only apparent when the coupling exceeds If we are using an impact discharger the double-period about 30 per cent, and die away with a very little reduction in the coupling, leaving the predominant free secondary oscillation as the survivor. These curves show how very quickly the primary spark is quenched when using the impact discharger. If the maximum secondary current is set up as ordinates in terms of the coupling as abscissa we obtain curves which rise up quickly to a maximum value value of the secondary current is determined both by the and fall again, and which indicate that the maximum coupling and the secondary decrements. The two previous papers were discussed together. Mr. A. CAMPBELL observed that the curves in the second paper showed the superiority of the impact method, and remarked that, since the losses at low frequencies were so exactly proportional to the square of the voltage, it was surprising the same was not true at high frequencies. Mr. DUDDELL also expressed surprise that with low voltages the losses were not proportional to V, and drew attention to the importance of knowing exactly where all the energy losses in the secondary occurred. Mr. E. H. RAYNER remarked that a more accurate way to test the results was to plot the power factor against the voltage. Mr. ADDENBROOKE said he thought the variation of the losses with temperature was very great, and should be taken into account. Dr. ECCLES pointed out that it was only after the primary spark was quenched that it was possible for the secondary circuit to emit its natural frequency. Prof. G. W. O. Howe exhibited oscillographs showing the oscillations in the primary and secondary circuits before and after the primary spark had died down. SOCIETY OF CHEMICAL INDUSTRY. (LONDON SECTION). Ordinary Meeting, February 6th, 1911. Mr. E. GRANT HOOPER in the Chair. Chemistry of the Lead Chamber Process." By F It was shown by Davy in 1812 that SO2 and the red vapour from N2O4 do not interact except in the presence of water. The formation of chamber crystals (nitrosulIm-phonic acid) is an intermediate step in this action, and their decomposition by an excess of water liberates a fresh quantity of nitrous gases. If in addition to sulphur dioxide and nitrous gases, oxygen (or air) and water be available, the conversion of the SO2 into sulphuric acid is complete. Chamber crystals, however, can never be observed when the chambers are working normally, and the various assumptions necessarily to Davy's theory are not true in In the course of the experiments described in the previous paper on the measurement of energy losses in condensers a large number of measurements had to be made with the cymometer of the frequency of oscillations in, and the inductance of, the secondary or condenser circuit. then an easy matter to draw complete resonance curves | practice. It was RÖNTGEN SOCIETY. The chamber gases, it is obvious from their colour, | abnormal conditions observed in the working of the contain more nitric oxide than is required by the relation chamber process, but by its application it was possible to NO + NO2. Finally, nitrosulphonic acid does not dissolve predict a pecuilarity of the chamber process, namely, the without decomposition in sulphuric acid as dilute as is possible formation of ammonia, of which no one had found in the Glover tower, but nitrosulphonic acid, a thought, or on the basis of the older theories could have reduction product of nitrosulphonic acid, is for some con- thought. siderable time stable in sulphuric acid of sp. gr. 16; i.e., acid containing 70 per cent of H2SO4. This reduction product imparts a characteristic blue colour to its solutions, and its copper salts are coloured an intense blue. From these observations we must draw the conclusion that nitrosulphonic acid, as an intermediate product in the formation of sulphuric acid by the chamber process, must be left out of consideration, inasmuch as the process continues in concentrations of sulphuric acid from 60 per cent at the end of the chambers, to 80 per cent in the Glover tower, and in the latter case the temperatures concerned are such that even a temporary formation of nitrosulphonic acid is out of the question. The assumption is almost forced on us that in nitrosisulphonic acid we have the true intermediate product of the chamber process. All experiments lead to the conclusion that when air reacts on nitric oxide, two different substances are successively formed. The first, which is produced after a very short time, dissolves in acids or in alkalis to nitrous acid; behaves, in fact, as if it were nitrous anhydride, N2O3. The second, which demands for its complete formation about 100 times as long, dissolves in acids and alkalis, half to nitrous and half to nitric acid; behaves as though it were N2O4, and reacts with water according to the equation N2O4 + H2O = HNO2 + HNO3. It must be admitted, then, that the nitrogen compound effective in the chambers is nitrous acid, HNO2, which exists dissolved in the tiny droplets of sulphuric acid which as mist fill the chamber; and the question of the chemical nature of the chamber process narrows itself to the inquiry, how, under chamber conditions, that is to say, in presence of air, water, and much sulphuric acid of 60 to 80 per cent, does sulphur dioxide react on nitrous acid? Is nitrosisulphonic acid really formed under these circumstances? The author shows experimentally, first, that sulphur dioxide does not react on nitrosulphonic acid; second, that sulphur dioxide reacts on nitrous acid, which is dissolved in sulphuric acid of the concentrations which occur in the chambers, with formation of nitrosisulphonic acid, which later breaks up into sulphuric acid and nitric oxide. Only one question remains to be answered, though that is an important one: How is nitrosisulphonic acid, H2NSO, formed from sulphur dioxide, SO2, and nitrous acid, HNO2? He also shows that when nitrosisulphonic acid is formed from sulphur dioxide and nitrous acid, nitric oxide gas is evolved. The remaining link of the chain is forged. The formation of sulphuric acid in the chambers occurs in this way, viz., the nitrosisulphonic acid formed dissociates into sulphuric acid and nitric oxide, and finally the nitric oxide, with air and water, becomes oxidised to nitrous acid. In the chemical sense, however, we are far from having finished. We ask further: What is the mechanism of the reaction by which, from 2 molcules of nitrous acid and I molecule of sulphur dioxide, nitrosisulphonic acid is formed? A chain of experimental evidence is adduced, which by a reduction of the processes going on in the chambers to well known and frequently observed properties of nitrogen compounds satisfies the last desire which the new theory of the chamber process still left to us. That theory can now be expressed in four equations, from the integral of which the catalyst disappears, and there remains an oxidation of sulphur dioxide by means of water and oxygen to sulphuric acid. 2NO+ H2O+0=2HNO2 HNO2+ SO2 = ONSO,H. + I. 2. 3. H2NSO = NO+H2SO4 · 4. SO2+O+ H2O = H2SO4 PROF. WERTHEIM SALOMONSON, of Amsterdam, was the lecturer at the meeting of the Rontgen Society on February 2nd. His subject was the Induction Coil, but he restricted himself to his own oscillographic researchusing the high-frequency pattern of the Duddell instrument-upon the behaviour of coils under varying working conditions. His oscillograms were taken with various well-known coils, and with at first a mercury break working in paraffin oil or alcohol, and afterwards with a mercury jet-interrupter working in gas. He devoted attention separately to the ascending and the descending portion of the curve of the primary current as described in the oscillogram. The form of the ascending part of the curve, which denoted the increasing current strength, was principally determined, he said, by the so-called time-constant of the circuit, the self-induction being divided by the resistance. The rapidity with which the curve rose depended upon the value of the electromotive force of the battery, divided by the self-induction. The descending part of the curve, representing the fall of the current to zero in about one-thousandth of a second at the moment the circuit was broken, was of greater interest, since it was in this phase that the secondary discharge occurred, giving the spark or lighting up the X-ray tube. In all oscillograms the rapid descent of the curve was to be noted, sometimes followed by a few oscillations. His oscillographic study of the descending curve had led him, he said, to the conclusion that if the capacity of the primary condenser were varied the time taken by the primary current to fall to zero varied also; that if the capacity were gradually increased, starting from a very small value, the stopping time of the primary current lessened until a definite capacity had been reached, and that any further increase of the capacity tended to lengthen the flow of the primary current after the commencement of the break. A considerable portion of Prof. Salomonson's paper was occupied with a study of the oscillations found in the descending part of the curve. These were due chiefly to the secondary sparks, and he showed that the amplitude of the more rapid vibrations depended upon the initial phase difference between the primary and secondary circuits, with a maximum for a phase angle of 90 degrees, and a minimum when there was no phase difference. The primary spark, in his opinion, only slightly affected the intensity of the oscillations. The number of interruptions per second for the maximum secondary effect was the subject which engaged Prof. Salomonson in the concluding part of his paper. After showing that with very rapid interruptions it was of advantage to reduce their number, and with very slow interruptions to make them more frequent, he pointed out that there must be some intermediate number which gave the best results from the point of view of secondary effect. Mathematical and practical observations had led him to the belief that for coils of 10-12 inch sparking distance the average number of interruptions to be desired was twelve to thirteen per second. Of course this number was only possible with very small voltages, say, up to 18 volts. If the voltages were increased there would be some danger to the coil in reducing the number of interruptions so greatly, and it would be safer to work with a larger number. For the higher voltages, therefore, he formulated a rule, which he admitted was somewhat crude and empirical, but which might serve for practical purposes, that the most advantageous number of interruptions Not only does the new theory account for various per second for small induction coils was given by the CHEMICAL NEWS, Feb. 17, 1911 Association of Teachers in Technical Institutions. number representing the voltage, granted that the timeeconomy of the interrupter were 0.5. With larger coils the number was smaller. Some radiographs of the thorax were shown which had received exposures of about one-hundredth of a second, a specially powerful coil, the "Unipols," being used for the purpose, and the definition was everything that could be desired. CORRESPONDENCE. 83 Technical Institutions) (1) The percentage of failure in chemistry has steadily increased from 244 per cent in 1907 to 469 per cent in 1909. (2) The percentage of failure in chemistry in 1909 and 1910 is disproportionately high as compared with the percentage of failures in the other subjects of the group. With respect to the above, the Council desires to point out that the educational experience goes to show that with a large number of candidates (about 700 in this case) from all parts of the country, the intellectual standard and training of the candidates, taken as a whole, will vary only to a slight extent from year to year, and therefore the ASSOCIATION OF TEACHERS IN TECHNICAL percentage of failures should show a corresponding minor INSTITUTIONS. variation only. Further, since the majority of the candidates take the same combination of subjects (Chemistry, Physics, Pure and Applied Mathematics) the high percentage of failures in chemistry, as compared with those in the other three subjects mentioned, cannot be justified. In this connection we would point out that there has been no alteration in the syllabus of the examination since about 1901. The Council desire to point out, with respect to the statistics relating to the External B.Sc. Pass Examination (Table II.), that in the period 1905 to 1909 the percentage of failures in chemistry is considerably higher than the TABLE I.-External Inter. B.Sc. 1906. Entries. 1907. 1908. Per cent 664 32.7 19'7 663 2315 1909. Per cent Entries. failure. 761 46'9 740 30.7 0000 734 495 15.8 Entries. Per cent 21.2 688 24'4 10'3 670 19'9 8.3 703 579 25'9 556 23.6 767 2513 14'4 180 ments of the examiners for a 'pass" in the respective | corresponding results in the subjects of Group A, namely, subjects, as shown in the great variation in the percentage Physics, Pure Mathematics, and Applied Mathematics. of passes. Further, the percentage of failures in chemistry has increased from 408, 1905, to 58.5, 1909, although (a) no alteration of syllabus has taken place, (b) it is generally admitted that the teaching of chemistry, along with that of other (scientific) subjects, has improved considerably during recent years.-I am, &c., P. ABBOTT, Hon. Secretary. I am directed by the Council to bring the following facts to your notice, and to ask you to place the matter before the Senate and the Council for External Students for their consideration. The figures in Table I., taken from the recent University Calendars, show that there is a serious discrepancy between the requirements of the various examiners, especially in the case of chemistry, and that as a result serious injustice is there done to students and teachers. The Council desires to suggest for your consideration, and that of the Senate, that the above statistics show :(a) That there has been an undue increase in the percentage (34'9 to 60-2) of total failures in the period 1906-1909 inclusive. (b) That in group A (the group of subjects taken by the great majority of the candidates, especially from the |