the hydrate. On rubbing it is more gritty than iron prepared as above. It is feebly magnetic.

Heated in hydrogen, ammonia is produced at about the same temperature as that at which the nitride is formed.

It readily burns in chlorine, ferric chloride and nitrogen being formed.

Heated in carbon monoxide, no evidence of the formation of cyanogen compounds could be obtained.

Steam at 100° slowly oxidises the nitride with evolution of ammonia. Hydrogen sulphide begins to react with it at 200°, forming ammonium sulphide and sulphide of iron.

The

Heated in nitrogen to the boiling-point of sulphur, no change occurs. temperature at which nitrogen is evolved by the action of heat alone must therefore be above this point.

An ethereal solution of iodine is without action upon the nitride.

From a slightly acidified solution of copper sulphate, nitride of iron deposits

copper.

Heated with ethyl iodide to 200° in a sealed tube, olefines are formed, and iodides of iron and ammonium, the reaction evidently being

5C2HI+Fe2N = 2FeI2 + NH ̧I + 5C2H ̧ + H.

Heated similarly to 200° with phenol no reaction occurred.

Treated with a mixture of hydrogen peroxide and sulphuric acid, analyses showed that very little, if any, of the nitrogen is oxidised, the whole dissolving as usual to form ammonium sulphate.

In conjunction with Mr. P. J. Hartog, the author has determined the heat of formation of the nitride by dissolving it in sulphuric acid contained in a platinum calorimeter, and observing the rise of temperature. Three well-agreeing experi

ments showed that the substance is formed with evolution of about three calories.

In general the nitride of iron behaves as an ammonia derivative, the nitrogen being either evolved in the free state, or converted into ammonium compounds, according to circumstances.

Its constitution may possibly be

5. Report on the Silent Discharge of Electricity in Oxygen and other Gases. See Reports, p. 439.

1. Report on the Action of Light upon Dyed Colours.-See Reports, p. 373.

2. Demonstration of the Preparation and Properties of Fluorine by Moissan's Method. By Dr. M. MESLANS.

3. Interim Report on the Formation of Haloils.

The Committee desired reappointment, as their work is unfinished.

4. Report on the Action of Light on the Hydracids of the Halogens in the Presence of Oxygen.-See Reports, p. 381.

5. On the Iodine Value of Sunlight in the High Alps. By Dr. S. RIDEAL.

During the past winter, at St. Moritz, in the Engadine, I had an opportunity of determining the intensity of the light as measured by the liberation of iodine from an acidulated solution of potassium iodide on the lines formulated by the Air Analysis Committee of Manchester. St. Moritz is at an altitude of about 7,000 feet above the sea level, and a succession of bright, sunny days can usually be relied upon, even in the depths of winter. The experiments in England, which have been carried out chiefly in towns, have not given a maximum value for the quantity of iodine that can be liberated by sunlight in one hour; and as the atmosphere in St. Moritz is not only free from haze, but is also remarkable for its exceptionable dryness, higher values than those likely to be obtained elsewhere were to be expected. Also, since the daily meteorological conditions of the place are carefully taken and recorded in the 'Alpine Post, the observations may possibly be of addi

tional value.

The instructions laid down by the Air Analysis Committee were carefully followed, and the solution of iodine taken as a standard was titrated with great care. The hyposulphite solution was checked against the standard iodine solution from time to time, and was kept in the dark when not in use. The only previous values obtained in Switzerland in the winter are those of Professor Oliver, who tells me that his winter average at Grindelwald for one hour's sunlight was represented by 16 c.c. of the thiosulphate solution, equal to 8 mgms. of iodine per

100 c.c. This number represents only the average of the brightest days, and larger results have been obtained in summer. My average for the nineteen brightest days in January of the present year is equal to 9.34 mgms. of iodine per 100 c.c. per hour. Owing to the situation of the village with regard to the surrounding mountains, the total amount of light per day is small compared with places which are less shut in; and, as will be seen from the accompanying table, the values given are for the hours during which there was a bright sunlight. The actual amount of possible sunlight on the days mentioned will be found in the meteorological records already referred to. It is interesting to note that in Manchester in January 1892, with a day of 83 hours' light, or nearly half as long again as at St. Moritz, the total light per week-day averaged only 4.5 mgms. iodine, or about that obtained during half an hour's exposure at St. Moritz. Even on comparing the Sunday values for the Manchester district, I find that the daily average is only 8.3 mgms., or less than the hour's average at St. Moritz. I believe that the comparatively large amount of sunlight per day experienced in the High Alps contributes largely in determining the hygienic value of a sojourn in these mountain health resorts.

The maximum hour value was 13.5 mgms. per 100 c.c. on January 1, and the lowest on January 24 of 5·7, and even this minimum was about 20 per cent. above the average daily value in Manchester.

6. On a Modified Form of Bunsen and Roscoe's Pendulum Actinometer.1 By Dr. ARTHUR RICHARDSON and J. QUICK.

In Bunsen and Roscoe's pendulum actinometer the oscillations of a pendulum cause a sliding shutter to pass backward and forward before sensitised paper, which is thus exposed for a known time and again shaded from the light.

In the present form an arrangement has been devised whereby the backward and forward motion of the shutter is brought about by a movement in one direction only.

This is done in the following manner: the shutter, which is made of a flexible material and in the form of an endless band, passes over the wooden rollers, the adjacent surfaces being brought close together by means of two additional smaller rollers. Two slits of equal length are cut in the band, so that when the latter rotates an aperture is uncovered when the slits overlap one another, and which again close when the band has travelled round a certain distance.

Beneath this aperture the sensitised paper is placed, which is thus exposed for definite times depending upon the length of the slits and the velocity of the band.

In order to bring about the movement of the shutter one of the rollers is connected with an eight-day clock, the escapement of which has been removed, the alterations in the speed, usually occurring when a clock is running down under such circumstances, being compensated by a fusee adjustment. Two advantages are claimed for this modification:

(1) It is portable, and measurements can be made when it is placed in any position.

(2) The time during which any portion of the slit is open (over the sensitised paper) is directly proportional to that occupied in opening the entire slit; since the rate at which the shutter moves is constant, whereas in the pendulum apparatus a series of calculations must be made to determine the length of time during which the slit is open for each mm. of its entire length,

7. On the Expansion of Chlorine Gas and Bromine Vapour under the Influence of Light. By Dr. ARTHUR RICHARDSON.

It was first observed by Budde that when chlorine is exposed to the influence of sunlight, an expansion of the gas occurs which is independent of the 1 Published in the Phil. Mag., xxxvi. (1893), pp. 459-463.

direct heating effects of the light. He also made a similar observation in the case of bromine vapour. These statements have been repeatedly called in question by other observers, who failed to obtain these results on repeating Budde's experiments.

Experiments made by the author, however, fully confirm Budde's results, and an arrangement is described in which the expansion of chlorine and bromine, as compared with that of air, under the influence of light, can be exhibited as a lecture experiment.

8. On the Cause of the Red Colouration of Phenol. By CHARLES A. KOHN, Ph.D., B.Sc., Lecturer on Organic Chemistry, University College, Liverpool.

The cause of the turning red of phenol has from time to time been the subject of investigation, but the published results are vague and conflicting. That even the purest carbolic acid of commerce becomes coloured on keeping has long been observed, and the general view of the cause of this colouration has been to trace it to some impurity or other contained in the phenol. By some the presence of a metal, especially copper or iron or their salts, has been regarded as the cause of the reddening, by others the colcuration has been attributed to alkalis or to cresol, which last in presence of the phenol has been oxidised with the formation of rosolic acid. Fabini, who more recently has investigated the subject, regards the colouration as due to the action of hydrogen peroxide on phenol containing metallic salts in presence of ammonia, the presence of all three reagents being necessary for the production of the colour.

Since oxidising agents, alkalis-especially ammonia-and metallic salts play an important part in the turning red of phenol, the action of these and similar reagents on phenol of varying degrees of purity was tried.

The phenol used was the purest commercial product known as 'absolute phenol, and in the later experiments a sample of specially pure phenol, kindly prepared by C. Lowe, Esq., of Manchester. The original product was repeatedly distilled from glass vessels and the distillates after one, six, nine, and sixteen distillations carefully tested with ammonia, hydrogen peroxide, caustic potash, mixtures of these reagents, and also with salts of iron and of copper both in the presence and absence of alkalis and of hydrogen peroxide. In all cases characteristic colourations ensue. That with strong ammonia is violet, and those with hydrogen peroxide, caustic potash, dilute ammonia, hydrogen peroxide in presence of caustic alkali, or of ammonia, metals or metallic salts with or without hydrogen peroxide, red or reddish brown. Each of the three reagents which, according to Fabini, must all be present in order to produce a colouration gives marked colourations on its own account. The blue colouration obtained with ammonia is identical with Phipson's phenol-blue,' and is probably phenol-quinone-imide. Sublimed phenol, as well as phenol prepared by the saponification and subsequent decomposition of gaultheria oil, behaves similarly.

Furthermore, all the samples thus prepared, and which were found on testing to be perfectly free from metallic impurities, turned red on exposure to ordinary moist air. Hence it is to be concluded that the purest phenol does redden of its own account, and not on account of the presence of impurities of any kind. This reddening does not take place in the dark, nor is it effected by the less refrangible rays of light. Phenol exposed in vacuo keeps colourless for months, as it also does when exposed in presence of water in absence of air, or in presence of air when perfectly dry. Both air and moisture are necessary for the colouration to ensue. It has been shown by Dr. Richardson that hydrogen peroxide is produced during the reddening, and to its formation the reddening of phenol when exposed to ordinary moist air is to be traced. The similarity of the colour produced by hydrogen peroxide with that which phenol assumes on exposure supports this statement. The colour is also produced by the electrolysis of phenol in acid solution. The colouring matter is not volatile, and the colouration is always accompanied by the absorption of moisture.

The nature of the colouring matter produced is still under investigation; the essential point so far established is that pure phenol possesses the intrinsic property of reddening when exposed to light in presence of air and moisture.

9. On the Rate of Evaporation of Bodies in Atmospheres of Different Densities. By Dr. R. D. PHOOKAN.

The results of his experiments showed that under the same conditions of heat and pressure a substance volatilises more quickly in an atmosphere of gas of lesser density than in one of greater. For instance, 0.05 grm. of naphthalin, heated in a bath of naphthalin vapour, volatilised in an atmosphere of hydrogen gas in 18 seconds, in air in 30, in carbon dioxide and nitrous oxide, both of which possess the same molecular weight, in 36 seconds.

Although these figures do not furnish sufficient data to determine the relative densities of the gases, yet they amply justify the above conclusion.

An atmosphere of vapour, on the contrary, seems to have no influence on the time taken for a substance to volatilise in it: 0·026 grm. of normal propyl alcohol, heated in a steam bath, took one and the same time, 12 to 13 seconds, to volatilise in vapours of such different densities as that of ether, methyl, and ethyl alcohol, chloroform, tetrachlor-methan, and ethyl iodide.

It is difficult to account for this anomaly. A certain difference in conditions in the employment of the two classes of bodies-i.e., the true gases and the vapours-must, however, be borne in mind, namely, that the gases were experimented with at a temperature much more removed from their point of condensation than that of the vapours.

It will be therefore interesting to know whether experiments made with vapours at a temperature equally removed from their point of condensation would not give results similar to those obtained from gases.

It may be that a vapour must attain a certain degree of energy or velocity of its molecules before it can act like true gases in influencing the volatilisation of a substance.

10. On the Occurrence of Cyano-nitride of Titanium in Ferro-manganese. By T. W. HOGG.

In this paper is given a short account of the fact there are probably about half a million isolated crystals of cyano-nitride of titanium in each cubic inch of the high percentage ferro-manganese now used for steel-making purposes, titanium carbide and nitride being also occasionally present.

The size of these crystals generally lies between 0.0001 and 0·001 of an inch, comparatively few of them being larger than this.

The number of crystals has been counted, and the lowest estimate gave 336,000 to the cubic inch of alloy; as a matter of interest, the weight of this number of cubes of cyano-nitride of titanium of 0.0001 of an inch has been calculated and found to be only 0-00003 of a gramme. Similarly, the weight of the same number of cubes of 0.001 of an inch weighs 03 gramme. The crystals are possessed of a high metallic lustre with brilliant mirror-like facets, and occur in the form of cubes, octahedra, and forms resembling the icositetrahedron; there are also present beautiful combinations of pyramids and prisms, and many of the cubes possess interesting symmetrical face modifications. As these different forms are all found together they are microscopic objects of great beauty and interest to the student of crystallography. These crystals are obtained by careful elutriation of the carbonaceous residue left after treating considerable quantities of the ferro-manganese with hydrochloric acid, cupric chloride, or dilute nitric acid: this latter is recommended as being the most convenient. In using it the mixture must be kept as cool as possible, and allowed to stand for about twenty-four hours; the larger crystals separate at once, the smaller forms being retained in the residue, which must be dried and gently pounded before submitting it to elutriation. This is best 1893. 3 A