Dec., 1868.

6 h. Insoluble silicates which, for want of a better name, I have in general called felspathic or clayey matter, appear to be almost always present in more or less quantity.

7th. The round masses so frequently encountered in quarries in our neighbourhood, and known as "Bastard whin," by whatever cause produced, appear to be simply parts in the sandstone rock which have become excessively hard from an increased proportion of carbonates, and are thus more or less easily separated from the surrounding rock.

8th. That the different red shades of colour observed in sandstones, more particularly in those belonging to the Devonian group, appeared to be due solely to peroxide of iron, and that the white rings and spots so frequently met with have resulted from the reduction and subsequent removal of the greater part of this iron.

ON THE COLOURED REACTIONS OF ANILINE, TOLUIDINE, AND PSEUDO-TOLUIDINE.

BY M. ROSENSTIEHL.

THE alkaloids used in the examination of these reactions were prepared with salts whose purity had been proved by the constancy-first, of the reactions, and secondly, of the solubilities, after fresh crystallisations. Since the first introduction of aniline into commerce, a large number of highly sensitive reactions have been made known, which were believed to be characteristic of this base; but the fact of the presence of pseudo-toluidine in this commercial product throws some doubt upon the worth of this means of analysis. By subjecting these reactions to a vigorous revision, I have discovered that there exists only one really characteristic of aniline viz., that discovered by Runge; this has been accused of being very variable, but a slight modification will not only render it much more certain, but also greatly increase its sensibility. If a few drops of a solution of chloride of lime be added to aniline suspended in water the intense blue colour first developed will quickly change to brown. In the presence of homologous alkaloids, the blue colour becomes gradually less apparent, disappearing before the brown products yielded by the toluidine; if, however, a little ether be added, and the mixture carefully stirred, all the brown matter will be retained by that solvent, and the water will assume a pure blue colour. Should the discovery of traces of aniline in a mixture (toluidine, for instance) be desired by this process, about I gramme of the alkaloid must be dissolved in 10 cabic centimetres of ether, an equal volume of water must then be added, and into this mixture a solution of chloride of lime, 1055 in density, is gradually dropped, carefully stirring after each addition. When only very small quantities of aniline are present the water gradually assumes a blue tint; and it is important to exhaust the action of the chloride of lime without introducing an excess of the reagent. According to experiments made in conjunction with M. Clemm, it appears that 5 cubic centimetres of chloride of lime, 1055 in density, are requsite for 1 gramme of alkaloid. The quantity of aniline contained in a mixture may be ascertained up to a certain point by working comparatively with a standard. By this approximate method, I discovered the presence of 2 per cent of aniline in M. Coupier's toluidine. In the above experiment pseudotoluidine may be substituted for aniline, in which case the water turns yellow, while the ether is simultaneously charged with a slightly coloured base, whose salts are of a fine violet red. If this etherised film be

311

decanted and stirred in slightly acidulated water the liquid will assume a tint comparable in beauty and intensity to that of a permanganate solution. This reaction is very sensitive, allows pseudo-toluidine to be discovered in the presence of the other two alkaloids, and would be still more clearly perceptible if the quantity of the former mingled with these latter were not so small. The reactions given by toluidine with chloride of lime are negative only.

Most of the other reactions proposed for aniline rests upon its transformation, by means of sundry oxidising agents, into Perkin's purple, which is well known to be transmutable by the aid of acids, into blue, green, and even yellow; these colours may be considered as corresponding with polyacid salts of mauveine. The blue colour being much the most intense, it is that which we should endeavour to induce, if the discovery of small quantities of colouring matter be desired; the dihydrated sulphuric acid, in which medium it is very above colour may be instantly obtained by means of stable, provided concentration of the acid remains unchanged.

All bodies which liberate chlorine or oxygen in the presence of sulphuric acid yields highly intense blue shades in connection with aniline and pseudo-toluidine; such are the chromates, oxygenated compounds of chlorine and manganese, binoxide of lead, chlorine, oxygen evolved at the positive pole of the pile, and an admixture of nitric and chlorhydric acids. Toluidine yields no colour with either of these reagents. Until now no means were known of discovering this base ; but if the reagent be changed, and nitric acid employed as an oxidising body, the actions are exactly reversed. When working in the cold, aniline and pseudo-toluidine give no colour, while toluidine yields a pure intense blue. This reaction is very delicate, and requires exactly the conditions described; dissolve the toluidine in dihydrated sulphuric acid, let it cool, pour several cubic centimetres of this solution into a thoroughly dry tube, and add a drop of nitric acid; the colour is produced in a second, lasts about a minute, and then changes to violet and red. This reaction offers the double advantage of allowing the discovery of small quantities of nitrates in the presence of chlorides and chlorates, and also of recognising small quantities of toluidine in a mixture, as, for instance, in commercial aniline; but then the colour produced is no longer blue, but varies from blood-red to blue-violet, passing through all the intermediate shades according to the proportion of toluidine; it is therefore important, in order to avoid error, to employ products exempt from chlorine. It is really surprising how little chlorine will suffice to turn auiline blue in the presence of nitric acid, and the colour, at first feeble, augments in intensity. This remarkable fact is easily understood, on reflecting that in the presence of nitric and sulphuric acids, of the concentration indicated, the chlorine must necessarily be indefinitely regenerated, and that by the same rule its action must be multiplied an hundred-fold.

The fact which I have described shows how delicate are the reactions, in consequence of their extreme sensibility, and that in order to avoid errors in observation, it is necessary to employ pure reagents.- Comptes Rendus.

DETECTION OF NITRATES IN WATERS.

BY THOMAS P. BLUNT.

THE following test for the presence of nitrates in a drinking water is a little more delicate than the com

mon one with ferrous sulphate; it depends on the reducing action exercised by sodium amalgam on nitric acid.

It was found that when a moderately concentrated solution of nitrate of potassium was poured over an amalgam of sodium containing about per cent of the metal, there was no evolution of hydrogen, ard on pouring off the supernatant fluid after some minutes, and applying the Nessler test, a considerable quantity of ammonia was detected. A solution of pure nitrate of potassium, containing 1 part in 1000, was then prepared, and measured quantities were added to different portions of per cent sodium-amalgam, consisting of about 200 grains each; it was thus found that 20 grain measures of nitrate solution, representing 1-50th grain of the salt, gave a just perceptible colouration with the Nessler test after standing over the amalgam for about twelve hours; 40 grain measure (= 1-25th grain salt) gave a marked reaction. Several attempts have been made to adapt the above test to the estimation of small quantities of nitric acid; they have at present, however, been unsuccessful, through the impossibility of pushing the reaction to completion; it is possible, however, that a system of comparative testing analogous to that at present adopted in the case of ammonia may lead to some results. As an example of the mode of applying the test qualitatively to a drinking water, the following account is given of an actual experiment:

To 2 ounces of a water known to contain a trace of nitrates, was added about 100 grains of a solution of hydrate of potassium, containing 1-12th of its weight of alkali; the whole was then evaporated nearly to dryness; thus any ammonia already existing in the water would be expelled. The residue was exhausted with distilled water which gave no reaction with the Nessler test, and the quantity of the solution made up to 200 grain measures, which were afterwards divided into two equal portions. One was at once tested with ferrous sulphate solution and sulphuric acid: a very faint brown colouration appeared at the point of junction of the layers of liquid, increasing considerably after a few hours.

The second portion was introduced into a carefully cleaned test-tube, with about 200 grains of per cent sodium amalgam: the tube was lightly corked to check diffusion of the ammonia formed as far as possible, but not so tightly as to prevent the egress of the hydrogen, which in dilute solutions of a nitrate is always evolved from the amalgam. The whole was left for about twelve hours: the liquid was then rinsed with successive portions of pure distilled water into a glass cylinder about 6 inches high and 1 inch wide, and the quantity was made up with distilled water to about 1000 grains. On adding about 15 grain measures of the Nessler test, a very strong colouration, accompanied by incipient precipitation, at once appeared. It is to be remarked that the liquid must always be decanted from the amalgam before applying the Nessler test, as the presence of nascent hydrogen appears to interfere with the action of the latter.

Shrewsbury, Oct. 12, 1868.

ON THE REDUCTION OF OXIDE OF COPPER TO THE STATE OF METAL BY MEANS OF INVERTED SUGAR.

BY M. A. COMMAILLE.

UNTIL the present time it was believed that the reducing action of sugar on salts of copper was arrested at

the protoxide (Cu.O), and that it was impossible to obtain the metal as such, when under similar circumstances a salt of bismuth was substituted for salts of copper, in which case metallic bismuth was precipitated. All the oxygen combined with the copper may, however, be removed by means of inverted sugar, but the metal is sometimes obtained pure, and sometimes as a mixture of copper and protoxide, according to the state of the liquid. Sometimes the copper presents the roseate appearance of galvanoplastic copper; at other times it will be much duller, but acquire brilliancy from polishing.

In order to obtain the reduced copper, take a very dilute solution of that metal and pour into it sufficient caustic potash to produce a precipitate; add to this liquid a solution of inverted sugar, and the precipitate will dissolve; the solution, which should not be acid, is then boiled; and after a short time a red deposit of protoxide is formed, which must be separated. The liquid is again boiled, and a fresh precipitate appears, which is proved to be formed of metallic copper and protoxide, which latter is removed by very weak chlorhydric acid; the undissolved precipitate is dried and polished with some hard body, when it presents the brilliant aspect of metal. On boiling the mother waters, by which the second precipitate was deposited, a third deposit is obtained, consisting only of metallic copper, and as red as galvanoplastic copper.

This metal may also be instantly obtained by the following method, without any mixture of protoxide:Before reducing the precipitate produced by the potash in the sulphate solution, by means of the inverted sugar, neutralise the acidity of the sugar solution, and when the precipitate of the copper hydrate is almost entirely redissolved, filter it, and boil the limpid liquid thus obtained; the metallic copper will then be seen to fall, but its colour is not quite so bright as in the former experiment.

I found it impossible, when working with M. Frommherz's liquid under ordinary circumstances, to obtain metallic copper; nevertheless it appears to me necessary to be on one's guard in saccharometry against the possibility of the complete reduction of oxide of copper.

ELECTRIC ILLUMINATION. SUNDRY EXPERIMENTS RELATIVE TO THE PRODUCTION OF ELECTRIC LIGHT.

BY F. P. LE ROUX.

THE brilliant phenomenon, now popularly known by the name of electric light, was first discovered by Davy; who, in 1813, conducted the current of a powerful pile between For more than half a century two pieces of charcoal. mankind has been dazzled by this light, whose brilliancy is comparable only to the sun, and yet many theoretical questions as to its intimate nature remain to be solved, and many practical difficulties with respect to its! extended employment to be removed. The experiments which form the subject of this article may be found interesting from both points of view.

On the Nature of the Voltaic Arc.-In order to examine in detail, and without fatigue, what takes place between the two pieces of charcoal in Davy's experiment, the phenomenon should be projected upon a screen, that is to say, an augmented image should be produced by means of a lens, an experiment which is now common. Two salient traits of this phenomenon M. Stolba obtained metallic copper by means of glucose.

will now be remarked; on the one hand, the vivid incandescence of the points of charcoal, which glow with the brilliancy of the sun; on the other, the soft bluish clearness of the intervening space.

This bluish space generally presents the appearance of a conical trumpet, whose larger extremity rests upon the positive, and its lesser upon the negative point of charcoal. The slightest breath will alter the shape of this bluish light, which curves from its centre in an opposite direction to the breath, while its extremities continue to rest on the most incandescent points of the charcoal; this has led to the conclusion that it is produced by a highly rarefied substance extending between the charcoal poles..

At the time these experiments were first made it was usual to arrange the charcoal points in a horizontal line, where the ascension of currents of heated air naturally produced the incurvation above described, and led to its receiving the name of the voltaic arc; this name it still preserves, although it is no longer justified by the external appearance of the phenomenon, owing to the vertical arrangement of the charcoal now general.

The voltaic are may be distorted or even broken by the proximity of a magnet; the sensible effect of a magnet on the arc is similar to that which it would have on a highly flexible circuit, permeated by the same current as the intervening space. It is evident that the voltaic arc is exactly the route taken by the electric current in passing from one point of charcoal to the other.

A glance at the whole phenomena will suffice to show that the quantity of light proceeding from the arc itself is minimum, and that the incandescent surface of the charcoal is the true source of the so-called electric light; but the portions of that surface, where the incandescence is incomparably more vivid than elsewhere, are the extremities of the sticks of charcoal, just at the places between which the arc extends.

What then is this arc? It is evidently composed of that constituent of the charcoal, which experience has shown to be conveyed from one point to the other. What form then does this matter take during its passage? Is the carbon merely disseminated as an excessively fine dust, or in a state of gas? This has not as yet been definitively answered, but some experiments which will presently be described, tend to cause the belief that carbon exists in the arc under the form of

vapour.

One circumstance which is immediately perceived on examining the pieces of charcoal is the great difference in their temperature. The positive charcoal is considerably more luminous than the other, and its incandescence extends over a longer duration. I am even inclined to believe that the negative charcoal is heated almost entirely by the radiation of heat from the positive charcoal on one hand and the arc on the other, and by the heat proceeding from the condensation of the matter conveyed by the latter. I have made one experiment of a very simple kind, but which, however, I have never heard described, which will show that the heating of the positive pole is owing to a special cause, the seat of which is the exact point where the voltaic arc joins the charcoal. The experiment consists in this:-The charcoal electrodes are first brought into contact in the ordinary manner, and then separated so as to produce a very short arc of only a small fraction of a millimetre, which is interrupted at the end of some seconds; the positive electrode will then be found to remain incandescent for some time, whilst VOL. III. No. 6.-Dec., 1868.

20

the extremity of the negative electrode will be scarcely red. By this mode of operation the heat evolved by the arc is almost entirely eliminated, as the latter from its shortness offers feeble resistance, and is consequently the seat of but a slight disengagement of caloric. If, on the other hand, the extremity of the positive charcoal be really the source of evolution of caloric, one would suppose that the caloric would communicate itself more rapidly by interior conduction to the mass of the positive charcoal than by radiation to the negative charcoal, and that if the passage of the current were interrupted for a few moments, a marked predominance of the first of these effects would be apparent; this, the experiment shows to be the case. There exists, then, at the positive pole, a special source of evolution of caloric hitherto unexplained, but which does not, to me, appear inexplicable; I shall, however, revert to this subject in a special publication.

Employment of a Current of Oxygen for Disposing the most Luminous Portions of the Charcoal to greater Advantage.-It has just been remarked that the greater proportion of light arises from those portions of the surface of the electrodes, between which the arc arises; that of the positive pole is slightly concave, that of the negative more or less convex; that of the positive pole exceeds the other both in size and brilliancy. As it is usual to arrange the electrodes in the same vertical line, the illuminating surfaces in question are necessarily presented obliquely with regard to a horizontal direction, which is generally that in which the light is used. If from defective homogeneity of the charcoal, or any other cause, these surfaces should bend in any direction, the illuminating power will be seen to vary in that direction also. With such charcoal as that produced by gas retorts, the only kind now obtainable in the present state of commerce, and very impure, the disturbance of the arc is perpetual, and in photometrical measurements, the illuminating power of an electric lamp is seen to vary at each moment within extensive limits. I believe that if the arc could be made to remain always inclined in a given direction—that is, towards the side on which the maximum of light is desired to be thrown-that the steadiness of the light would be secured; this was satisfactorily proved to be the case; the arc being forced to incurve, the surfaces which formed its base were so much the more vertically inclined, and the light thereby more clearly directed towards the region desired to be illuminated. It is particularly necessary that the current of gas should incline the electrodes in a different direction to that in which the arc is bent, otherwise they will approach each other too closely, and the arc, choosing the most direct road, will also change its position. It is exactly this combustion of the charcoals, which is performed by a current of gas, which turns them towards two eccentric points, between which the arc naturally maintains itself so easily that the slightest force is sufficient to incurve it as we have described. The gas pipe should be placed at about 1 centimetre from the char-coal electrodes, and slightly below a horizontal plane passed between them, in order to combat the tendency acquired by the gas as it heats to flow in greater quartity towards the upper electrode. A good result is obtained by causing the gas to escape by two holes of about 1 millimetre in diameter, and 1 millimetre apart; the consumption of oxygen is not great, and may be estimated at about 15 litres per hour. It is of course understood that the conditions must vary with the size

of the electrodes and the force of the electric current employed, the stream of gas being weaker and weaker in proportion to the diminution in the size of the electrodes.

I should recommend all who desire to put this experiment into practice to carefully isolate the tube containing the gaseous current, as without this precaution the are might come in contact with it and cause its instantaneous destruction.

Arrangement for Subjecting the Voltaic Arc to Spectral Analysis.—The image of the voltaic arc and luminous electrodes being already thrown upon a screen, let us conceive a small opening to be made in that part of the screen upon which the arc is projected, and thus an extremely simple means is obtained of isolating the light emitted by the arc from that of the electrodes. By placing the slit of a spectroscope behind this opening the character of the light emitted by gaseous substances is instantly perceived-namely, discontinuity. The spectrum of the voltaic arc is perhaps the most beautiful spectacle offered by the spectroscope; it appears to me to resemble most the spectra observed by MM. Plucker and H ttorff* during the combustion of cyanogen in oxygen, and also by M. Morren in that of carbon.t

I intend to make an exact determination of this spectrum as soon as it is possible to use a purer carbon than the charcoal of gas retorts; at present we can do no less than infer the presence of gaseous carbon in the voltaic arc. The proceeding just described may perhaps be also applicable to the examination of different portions of other kinds of flame; it will be only necessary to place a screen at the distance of about a few millimetres from the slit of the spectroscope, and to there project by means of a lens, the image of the flame to be exarained; by this means it may be ascertained at any moment what part of the flame is under the spectral analysis.

Spontaneous Reappearance of the Voltaic Arc after an Interruption of Short Duration.-About twenty years ago, M. de la Rive, when studying the effect of magnetism on the voltaic arc, succeeded in producing it between bars of soft iron; he then observed that under certain conditions the arc was sometimes broken by the influence of an electric current circulating around these bars, but that if the disturbing influence was removed before the poles had cooled, the arc sometimes reappeared without obliging him to bring the poles into contact.

Ten years later, M. Wartmann, of Geneva, in experimenting on the practical application of electric light, proved that "it was possible to suspend the circulation of electricity between two charcoal electrodes during 1-20th of a second, without, he says, the arc disappearing.

M. Wartmann considered that, for a certain time after the interruption of the current, certain solid particles remained interpolated between the two poles. I was led to a inore detailed study of these phenomena, by endeavouring to give some description of the production of electric light by means of magneto-electric machines, in which case the current changes, as is well known, many times in a minute, and must therefore of necessity pass by zero-that is, it must be interrupted.

Plucker and Hittorff. Philosophical Transactions, 1865. A. Morren, "On the Flame of some Carburetted Gases." (Annales de Chimie et de Physique, 4th series, vol. 4, 1865. Bibliothèque Genive (Archives des Sciences Physiques et Naturelles, vol. 36, page 325, 1857-)

The first question to be asked is, Is the arc really interrupted at the same time as the current? Observation has proved that the arc really ceases to exist during the interruption of the current; the following is a convenient method of demonstrating it to a numerous audience.

At the end of a pendulum, beating about every half second, is suspended a screen upon which the reflection of the electrodes is thrown by means of a lens; by this contrivance the reflection of the electrodes is displaced upon the screen. Consequent upon this is the displacement of the image formed upon the retina, whose sensation is thus rendered less obtuse than it would be were the image continually in the same place, and the changes of short duration which the object under examination may undergo are more eas ly perceived.

These arrangements being finished the two electrodes are placed at a distance from each other of about 2 millimetres, and there fixed (if they be mounted on a regulator the movement must be stopped), and the current is then interrupted by the hand for a period not exceeding 1-20th of a second. This interruption may be more easily made by a spring plate working against a catch, ti.e arrangement being such that the separation may be made by a downward movement. It will be clearly perceived that the violet light constituting the arc disappears at the moment when the current is interrupted.

This interruption may be repeated a large number of times per second (we shall presently recur to the use made of this fact in the division of electric light) so that the electrodes may be examined during the length of the interruptions only, by an apparatus somewhat resembling the phosphoroscope of M. Edm. Becquerel; the interruptions are effected by the apparatus itself, and thus the condition of the intervening space may also be examined during the same period. Economical motives have hindered me from realising this project, and I may here add that the kindness of M. Serrin first enabled me to effect the experiments which form the subject of this communication.

On a Mode of Dividing Electric Light.-The fact here admitted as the result of experience that, in order to produce electric light most advantageously, it is necessary to employ a considerable number of eleinen's arranged in series, may be of some theoretic consideration; operators usually employ fifty elements of Bunsen. Light may be produced by a much smaller number-twenty, for instance, but the quantity of light produced is proportionately less, and the working condition arising from length of arc, regular consumption of charcoal, and exhaustion of pile, is much less favourable than when a larger number of elements are used.

On the other hand it is an advantage in many cases to be able to distribute the entirety of light yielded by a pile or electro-magnetic machine into several distinct lamps.

The only means which have been hitherto tried to my knowledge are, after dividing the source, the introduction of several arrangements into the same circuit, and the bifurcation of a single current into several lamps.

Thus, as to the division of light into two burners, and supposing the source to be a pile of fifty elements, either of the three following dispositions may be made.

Ist method. Dividing the source.-Compose two piles of twenty-five elements each, and conduct each to a special electric burner.

2nd. By succession of burners.-Dispose the fifty ele

ments in tension, and conduct the same current through the two arrangements successively.

3rd. Bifurcation.-Retain the fifty elements in tension and direct the current into two arrangements. We have already stated that the first system of two small piles was less advantageous with regard to the entirety of light than the ordinary system of one.

The second system has no advantage over the first, but possesses the disadvantage that if by any means the condition of one of the burners be altered, the other will sympathise; therefore, as from the very nature of all regulating apparatus hitherto known, the movement of the charcoal depends on the intensity of the current; movements induc'd by the state of the electrode in one regulator, will indubitably occur in the other if they be really equal. If one were more sensitive than the other it would be always separate, and consequently there would sometimes be a contact of the electrodes, and sometimes a rupture of the arc from too great elongation.

The third system by bifurcation or derivation has not the same drawbacks; the two arrangements do not influence each other in the same inconvenient way, but fall under the same conditions as to intensity as in the method of equal division of the source.

I have meditated dividing the current into periods, in order to separate the light into spaces, profiting by the power possessed by the arc of spontaneous re-establishment after an interruption of short duration.

Let us suppose that by means of appropriate mechanism the current is made to pass into one arrangement during 1-100th of a second, theu during the next I-100th into another arrangement, and again back to the first during the next 1-100th, and so on. As the duration of the interruption is short, the arc re-establishes itself spontaneously; and as it is also weaker than the permanency of the impression received by the retina, the arc itself being but a feeble source of light, the eye will consequently receive a sensation of continuity; moreover the incandescence of the electrodes varies only slightly. It is not to be denied that this experiment presents more theoretical than practical interest on account of the complication involved by the mechanism necessary to divert the current into sundry arrangements, nevertheless in some cases it may be found useful. The chief difficulty in applying it arises from the destructive action upon the distributive wheel of the sparks which arise in consequence of vibrations which it is impossible to quite avoid; the contact separates from the surface of the wheel; these sparks, which are so many voltaic arcs formed at the expense of the opposing surfaces, corrode and render them unequal, and aggravate the causes of interruption.*

If the apparatus works regularly the sparks will be feeble, but they will be altogether fatal if one of the burners should act alone, and this vexatious occurrence should be above all avoided, which may be easily done by arranging in such a way that if by any means one of the burners should be extinguished, the current should still find an outlet on that side.

Experience has proved, in contradiction to a generally believed opinion, that it is better not to lubricate with any liquid the friction produced by the contact with the wheel; greasy bodies inflame at each spark, and thus augment its destructive effects by means of the developed heat; moreover, wear is not completely avoided, and the grease agglomerates the detached metallic particles in an inconvenient In the dry friction the wear is more rapid, but it is regular, and the metallic dust is carried off by the disturbance of air; the general rule against the friction of two identical metals against each other must of course be obeyed, and the red copper contacts will succeed very well with the bronze wheel.

manner.

To render the sparks as harmless as possible, it is a good plan to multiply the contacts, it being apparent, that if instead of one spring we substitute two or more, there will be some chance that at least one may be always in perfect contact with the wheel. This, however, will not be the case if a cause of displacement, common to all, should arise, which is exactly what happens at every change in the teeth; for this reason only a small distance, not more than a millimetre, should be left between any two consecutive teeth, and the contacts should be arranged so as not to leave one tooth until they have slightly touched upon the next.

The surface of the contacts being more easily repaired than that of the wheel, it is preferable to cause communication between the former and the positive pole.

The vibrations of the contacts may be lessened with some success by mounting them upon springs composed of several thin plates, alternated with layers of guttapercha.*

more

or

Combination of the Incandescence of earthy Oxides with that of the Charcoal Points between which the Volatile Arc is Produced. In applying electric light the method generally proposed is to direct it into a less limited region of space; all which escapes into the opposite region would be lost if it were not collected by reflectors more or less appropriate to the purpose. On the other hand, experience has proved that the voltaic arc is prone to irregular displacements, consequent upon inequalities in the cohesion of the charcoal, impurities contained in it, and above all the slightest agitation of the air. The most luminous portions of the charcoal electrodes being, as we have already remarked, the surfaces between which the arc arises, these surfaces are inclined sometimes in one direction and sometimes in another by reason of the displacement which the arc undergoes, the result being a considerable variation in the effect of light produced by the latter in any determinate region.

I believe that if there could be placed on the opposite side to that towards which the light was to be directed, and in proximity to the arc, some body capable of reflecting back in a luminous form the enormous number of radiations thrown upon it by the electrodes and the arc itself, these radiations would be more profitably utilised than by any other method, the arc being at the same time protected by a sort of screen, annulling in an almost hemispheric region all the abovementioned disturbing causes.

The substance chosen for such a purpose should be at the same time a bad conductor of heat, and possessed of great powers of radiation, conditions fulfilled to a great extent by I'me, magnesia, and earthy oxides in general. I first effected the experiment with cylinders of magnesia compressed according to the process of M. Caron, and manufactured for purposes of oxyhydric illumination. By placing the base of one of these cylinders, whose diameter is about 8 millimetres, at a short distance from the charcoal points of an electric lamp, in such a way that the magnesia may be, as it were, licked up by the voltaic arc, it will assume an

These considerations of course apply to cases where a pile is used as the source of electricity; if a magneto-electric machine be employed, it must be arranged in a peculiar manner. The mean tension of the machines usually employed is known to be so weak as to produce only an arc of short extent, and as experience shows that the bifurcating system described necessitates a reduction in the length of the arc, the machines should be so managed as to present a higher degree of tension; it will be as well to arrange the distribution in such a manner that it should occur at the moment when the current is at the minimum in order to avoid the intensity of the sparks.