Faraday report to Trinity House   29 October 1838

To the Master, Deputy Master¹, and Elder Brethren of the Trinity Board.

My Lords and Gentlemen,

I have the honor to report for the fourth time upon certain Lamps invented by Mr. Goldsworthy Gurney, and placed before me by your directions; and I have also the very high pleasure of announcing that amongst them is one, which I unhesitatingly recommend as likely to be of very great service in the illumination of Light Houses. Since my last report² I have had three different lamps under trial. I feel myself bound, according to my instructions, to report upon each of these; but the two first become unimportant when put in comparison, (in their present state), with the last or Good Lamp.

Vapour or Gas Lamp, No.4.

This lamp, which I received on the 7th of April of this year (1838) was constructed on the principle of converting the oil, or other fluid fuel to be used, into a mixture of gas and vapour, by the action of heat upon it whilst passing through a small retort. When thus volatilized, the fuel was sent into a cylindrical chamber, representing the burner of the large lamp, described in the last report; and when lighted and burnt as it issued forth formed a large cylindrical flame; oxygen was then thrown into the middle of this flame, and of course highly exalted it’s combustion. If the sketch in the last report be referred to, the parts there marked oxygen and oil, exactly represent the places of the oxygen and gaseous fuel in the present lamp.

The oil gas retort was (in the present case) an iron tube about 18 inches long and 2 inches in diameter. It was placed in a small portable furnace so as to be raised to a red heat. The oil was allowed to fall drop by drop into a funnel connected by an S tube with one end of this retort, and the gas, when formed, was carried forward by a tube connecting the other end of the retort with the burner of the lamp.

This lamp burnt pretty well upon the day when I saw it. That was the only occasion upon which I had it in use; and no measurements of its power, or consumption of oil and oxygen, were made. The jets charred; the supply of oil vapour was irregular to the different parts of the annular flame; and a little explosion occurred in the oxygen chamber beneath the jets. To obviate these and other objections it required various alterations and adjustments; which are now probably rendered unnecessary, by the production of the Good Lamp No. 6 presently to be described.

Refrigerated Annular Lamp, No. 5.

This lamp which I received on the 31st of July of the present year (1838) was intended by it’s construction to meet the objections that were, or might be, raised to that described in the third report. It has, like that, two cylindrical cottons, an inner and an outer, of the respective diameters of 2¾ and 4⅛ inches; but these, instead of being fixed to wire gauze, have a few fine iron wires wove into them, along and across, to give the requisite stiffness.

The oxygen enters the flame in the same direction as before; i.e. from apertures arranged in a ring 3⅝ inches in diameter; but these apertures are only 18 in number, (instead of 40) and they are the terminations of so many upright independent jets, 0.4 of an inch in height; which being equidistant, are fixed into one common annular chamber by means of which the oxygen gas, issuing from them, is supplied.

In the former lamp the parts near the flame became so hot, that, as described in the last report, the oil which flowed over the outside of the burner would take fire, and the combustion would extend ¾ of an inch or more downwards over the metal. This effect not only wasted oil but it reddened the flame, and also caused those little carbon cones to form about the oxygen apertures which are found so greatly to diminish the light of the lamp. To prevent such results with the present lamp, Mr. Gurney has adopted a plan of keeping the oil cool, from which I have provisionally named the lamp. He has placed a cylindrical chamber beneath the oxygen chamber, in the midst of the oil in the burner, a small pipe brings water from a tub into this chamber below, another pipe takes the water off from above; by allowing a small stream of water to pass in this way through the middle of the oil in the burner it was found to preserve it, as a mass, in a comparatively cool state, and the evils consequent upon it’s over-heated condition were entirely removed. No practical objection arose to the adoption of this plan. Four gallons of water were enough for two hours, and that which had run through the lamp was ready at the end of that time, to be returned for a second cooling.

I have spoken several times of certain little carbonaceous cones which tend to form at the summits of the oxygen apertures. These, by their shape, disturb the course of the oxygen into the flame, mix it with carbonic acid, and greatly reduce it’s power of evolving light. To remove them, Mr. Gurney has attached a little jack to the lamp; the hot air in it’s ascent causes the jack to revolve, and the latter carries a tripper or finger, the end of which passing just over the aperture, removes, in its revolutions, any carbon cones that may be on them. This expedient was found very effectual in practice.

This lamp had no glass round its flame, but it being essential that the flame should not draw in and collapse much, as it ascended, as then oxygen would be blown through uncombined and wasted, Mr. Gurney applied a hemisphere of copper under the jack, with the convex part downwards; this had the effect of opening out the flame and making it assume a shape, which was nearer to what it out to have, namely, that of a cylinder.

<-><-><->

Such was the construction of the lamp; I will now describe it’s general condition when burning. The flame was very smoky, and much carbon rose from it and passed away through the funnel pipe. This indicates a waste and loss of fuel; and though the effect could be prevented by turning on more oxygen, it would only be, by involving a loss of that gas, which would be equally against the economy of the lamp. The flame was about 3 inches high, but by far the greatest portion of light came from the lower part of it: for, on screening off the upper part and leaving only from 1¼ to 1½ inches of the lower part visible, the light appeared to be very little diminished.

The iron wires in the cotton, by not burning away or contracting with the rest of the wicks, gradually formed projecting parts on which, large lumps of charcoal would collect, and by their bulk and interference with the regular access of the air, destroy that uniformity of shape which it is important the flame should possess: for whenever the flame is irregular, either oxygen is wasting in one part or fuel in another. The jack belonging to the tripper was also liable to become encrusted with soot.

The lamp burnt with considerable steadiness for two hours at a time; after which, from the charring of the cottons and the consequent diminution in the supply of fuel, the light decreased. On such occasions the cottons required trimming; this was easily performed within one minute of time and might be done frequently in a shorter period, since it is the outer cotton principally which thus chars. The trimming of the lamp during combustion appeared to be very easy and to require no suspension of the process of burning; both the inner and outer wick being easily accessible by properly constructed scizzars.

<-><-><->

I now proceed to describe the results as to the amount of light and expence. The trials were made on three different days, and the average results of each day I will give in the form of a table. The proportions of light were purposely varied on the different days for the purpose of ascertaining how far the lamp was manageable in that respect; and what the variation of expence, consequent upon such variation in the light, would amount to.

The average of the three days trials, gives the price of light equal to that of a Standard Argand lamp for 12 hours, when obtained by this lamp, as 15.3 pence

To these I may add from the former reports

These results show the lamp to be one, which, being manageable in it’s practise, will give a light ranging at pleasure from a quantity equal to 20 or fewer Argands up to 32 or more, and in that respect able to evolve from the same focus, three times the light of the French lamp. It’s cost is also less, light for light, than that of the French lamp; but is more than that of the Quintuple or first large lamp. If the present lamp had a glass, the flame would be much steadier, and I have no doubt the expence would be less, because of the better form which the flame would then assume, and the more effective use which would consequently result both of the oil and oxygen.

<-><-><->

It will be very evident that the three last lamps have not been improvements upon the first great or Trinity Lamp. Mr. Gurney therefore varied in a certain degree his mode of bringing the oxygen and fuel into relation, and by returning towards, and even passing beyond, the principle of separation of flame, which in the first lamp had given such good results, has at last surpassed all his former efforts, and produced an arrangement which I think cannot fail of being extensively useful.

I received this lamp (No. 6) on the 2nd. of the present month. I will distinguish it provisionally by the title of the Good Lamp; and will first describe it’s construction. A small circular cotton nearly half an inch in diameter, is contained in an oil tube half an inch in diameter. Up the middle of the tube rises an oxygen pipe surmounted by a jet, the termination of which is 5/16 of an inch above the level of the oil tube. Thus arranged, if the cotton is lighted, a flame like that of a coarse round wick is obtained, no air having access to the interior of it; but when required, oxygen can be introduced through the jet into this internal part, and being so admitted the intensity of the light is exalted to a very high degree.

Seventeen of these constituent lamps are arranged in a ring; all the oil tubes being fixed on to one common oil chamber, and all the oxygen tubes being also in communication with one common chamber of gas. The oil tubes are seven inches in length and 1/10 to ⅛ of an inch apart from each other, so that the common air can pass up regularly between and around each tube; the diameter of the circle formed by the 17 oxygen apertures is 3⅝ inches.

A glass is used with this lamp and produces very excellent and important effects. It renders the flame of each constituent lamp vertical, so that the oil and the oxygen act upon each other in the most effective manner. It protects the flames from external currents of air, and renders them very steady; and it also, by causing a rapid ascent of cold air between the oil tubes, cools them so much, that there is no occasion for the refrigerating water apparatus used with the last lamp. Mr. Gurney has placed a jacket of wire gauze about the lower part of the oil tubes to assist in cooling them by radiation; but I do not think it produces any sensible effect and would probably be better away.

The carbon cones form so rarely in this lamp that there is no occasion for a tripper or finger, and consequently, it, with its revolving jack is removed. The glass also is so effectual in giving the right form to the flame that there is no occasion for the hemisphere in use with the former lamp. The cottons are fixed on to little holders, themselves attached to wires which pass downwards through collars in the bottom of the oil vessel, so that each can be readily adjusted by itself. The lamp is, in it’s construction, as simple as any former lamp; and in practice, very manageable. The oil is fed, as in all these lamps, by hydrostatic pressure, according to the plan introduced by the Deputy Master with the Menai French Lamp.

When this lamp is burning, the flame is exceedingly intense, and steady; it is not smoky; and the glass keeps perfectly clean for hours together. The compound flame is four inches in horizontal extent and three inches in height; but, the important and effective part of the flame is not more than from 1¼ to 1½ inches high when giving the light of 25 or 30 Argands.

When lighted and adjusted the lamp alters for the first quarter of an hour; and on that account requires a little attention. If oil enough be supplied by the wicks at first it soon increases in quantity so as to require diminution. The cause of the change is simple, and depends principally upon the heat communicated to the oil tubes in the first 10 or 15 minutes; part of it’s effect is to cause a more rapid vaporization; another part is the heating and expansion of the oil within the tubes, which becoming lighter, stands at a higher level than before; and the remedy for both is, either to depress the supplying oil vessel a small quantity, or to lower and readjust the wicks. After the first quarter of an hour the tubes attain a steady temperature and no further change of this kind takes place.

The cottons burn very well for two hours; at the end of that time the tops are generally charred so as to require recutting, but the parts below are not at all hurt. They are easily trimmed, a single cut across with a pair of scizzars being generally sufficient for each. It would be easy by a properly shaped cutting instrument to trim them without removing the Glass; so that the action of the lamp need not be suspended for this operation. I do not think it would produce any effect upon the light sensible to the Mariner.

Carbon cones form very rarely and are easily removed, when they do occur, by the attendant.

In order to illustrate the illuminating power of the lamp and the cost of it’s light, I will arrange the results of experiments made on three different days in a tabular form, so as to give data corresponding to those already given in this report in relation to the Refrigerated Annular lamp No. 5 (page 61).

(from former statements)

The average of the three days gives the price of light

Here then is a lamp, more practical and convenient than any of the former lamps; giving light up to the high amount of 40 Argand burners or four times that of the French lamp, and still higher if needful; more economical than any of Mr. Gurney’s former lamps; more economical light for light than the French lamp in the ratio of 100 to 174; and apparently supplying light even cheaper than a common Argand lamp.

<-><-><->

But having attained to this condition of things, I was anxious before forming a final conclusion respecting this lamp, and it’s relation to the Argand and French lamps in light and expence, to re-examine the consumption of oil of the two latter. The Argand light has been used as the general standard; but without special care it is not possible to have the lamp burning with the same intensity at different times; and therefore I made some final experiments for the purpose of greater accuracy in this respect. The French lamp also I have heretofore taken, both as to light and expence, upon the report of Lieut. Drummond; and accordingly thought fit to re-examine and compare it afresh.

The Argand standard lamp was brought into good condition, and it’s flame, not only measured but drawn on a scale, so that at other times it could be adjusted until it possessed as nearly as possible the same power. I then ascertained, for myself, how much oil was burnt by it in 12 hours, keeping the flame during the whole time at the standard height; and found it to be 0.8 of a pint of oil, which at 6s/10d per gallon would cost 8.2 pence.

Gurney’s Good Lamp on two occasions, namely the 8th and 18th of October, this and the Argand lamp in the state just described, were compared; the mean result was that the light equal to one Argand (consuming the given proportion of oil) for 12 hours when obtained by the Good Lamp cost 10.42 pence. (page 65.)

French lamp. This lamp being put into action & lighted was brought into what I and Mr. Hall³ supposed to be a very good condition; after which attention was given to keep it in that state. Whilst so burning, it was found to give a light equal to 9.68 Argands supplied with the given proportion of oil, namely 0.8 of a pint in 12 hours. The consumption of oil to produce this effect in the French lamp was found to be at the rate of 14.4 pints per 12 hours. Thus the light of 9.68 Argands for 12 hours costs by it 12s/3½d; or the light of One Argand for the same time 15.25 pence.

According to these results therefore light

the three being as closely compared as they could be in so few experiments.

<-><-><->

Whilst burning with these proportions of oil, and oxygen, I compared the Good Gurney, and French lamps together, in their power of illumination, and found the former to be as 2.52, the latter being unity or 1. This was on the 18th of October when the former was burning only 0.92 of a pint of oil per hour and the latter was consuming as much as 1.20 pints in the same time. The Gurney lamp was very steady in its light; but the French lamp I found liable to frequent variations, and often requiring adjustment to prevent the diminution of the light or the smoking of the flame.

<-><-><->

Whether I consider the general comparative results obtained during all the trials, and referred to in the various reports given in; or whether I take only those in which I more strictly compared the Argand, French, and Good Gurney lamp; still I am bound to recommend the latter to the favourable consideration of the Honorable Board. The two great points upon which I base this recommendation are, first the price of the light obtained by the use of the lamp, it not much exceeding the price of light obtained by the use of the common Argand lamp; and second, the facility with which a steady, regular intense light can be obtained and preserved in order for many hours together.

There are also other points of recommendation; and what may prove a great one, is the variation which the construction of the lamp allows, consisting, as it does, of constituent burners, these may be used singly or in association, and so a light procured at the same relative price, but varying from that of one Argand and a half to forty or fifty; the highest proportion being compressed into a space not more than 3½ inches in diameter. This facility of variation gives the power of using the light either with reflectors or refractors at pleasure; and, combined with the relative price of the light must be of great importance.

Unless I have made some great mistake the advantage, as compared with the French lamp, is very great. The combustion of oil in the French lamp is, as might be expected, very different at different times; but of course it’s light will be different also if the oil in all cases be equally well burnt. Taking the returns from the lamp at the Menai for the month of July 1838, the average consumption is at the rate of 10.524 pints for 12 hours. The returns for the same month for the lamp at the Start Point is 12.24 pints per 12 hours. My result when I burnt the lamp at it’s best and endeavoured to compare it strictly with an Argand was 14.4 pints per 12 hours; it then gave the light of about 9.68 Argands consuming each 0.8 of a pint per 12 hours, i.e. it consumed 18 times the oil and produced not quite 10 times the light. Lieut. Drummond gives the consumption of the French lamp still higher namely 16.6 pints of oil in 12 hours, it then producing the light of 9.75 Argands burning 1.0625 pints in the same time so that the oil being sixteenfold the light was only 9¾ fold. It is not therefore merely from the direct comparison of the French with the Good Gurney light, but the comparison of both these with Argand lamps, burnt by myself and others at various times, that the following conclusion may be drawn; namely, that for the same money, the Good Lamp will produce from one and a half, to almost twice as much light as the French lamp in the same, or even in a less space.

Added to the previous recommendations of the Good lamp are those which may fairly be drawn from expectations of improvement in the preparation of the oxygen, and the use of cheaper oils or fuels. I have in a former report expressed my strong opinion that when the oxygen is made daily, the wear and tear of the retort will be less, and the consumption of fuel much diminished, I also think it probable that the Manganese used will still find a price, though a low one, in the market; and that these and other circumstances will importantly diminish the cost of the oxygen, or that substance, the use of which in the good lamp makes two thirds of the whole expence. With respect to the lamp-fuel it is exceedingly likely that with the employment of the tripper or finger already described in the lamp No. 5 to remove such carbon cones as may be formed, cheaper fluid fuels than the oil as yet used may be burnt; and thus a diminution of expence be effected in that direction also.

<-><-><->

I shall take the liberty of introducing into this report a few familiar remarks upon the causes and principles on which the power and economy of the recommended good lamp depend.

The same quantity of carbonated fuel, when burnt, may be made to give either a small proportion of light or a much larger one, the light not being necessarily in any constant proportion to the fuel employed; and it is upon this variation that the good and profitable use of the Gurney Lamp depends. An instance of the kind may frequently be seen in the streets, when gas illuminations occur on a windy night, for at one moment the light of the flame will be very great and at the next scarcely visible though in both cases the same quantity of gas is being burnt in the same time from the same burner. The like effect is produced, though in a smaller degree, with oil in an Argand Lamp, when burnt with or without it’s glass; or when burnt with a bright or a smoky flame.

Many of the results obtained in comparing the lamps are very interesting in illustration of this effect. Thus, with the French and the Gurney lamps, (page 66) whilst the latter was burning only 11.04 pints of oil in 12 hours, it gave 2.52 times the light of the French lamp burning 14.4 pints in the same time; so that the same quantity of oil which in the French Lamp produced a light of 100 could in the Gurney lamp produce a light of 328.

This striking difference in the light produced, is easily comprehended, when it is considered that the light is not any thing which must of necessity be evolved in proportion to the extent of chemical combination which takes place, but is almost entirely the consequence of the ignition of carbonaceous particles, which, as Sir Humphrey [sic] Davy has beautifully shewn, are for the time, present abundantly in the flame⁴. These particles, being ignited, glow, as a lime ball if similarly ignited would do, and it is their higher or lower ignition by the heat evolved from a concentrated or a dilute combustion, which is the principle cause of the increased or diminished light.

It is by thus making the fuel produce a far higher amount of light duty than before, that a portion of it can be spared, and it’s value be appropriated to defray the expence of the oxygen necessary to produce the requisite intensity of combustion; but, it is certainly a surprising, and to me unexpected, thing to see two thirds of the cost of fuel for a given portion of light suddenly suppressed, and the amount expended in the supply of a supporter of combustion, which before cost nothing; and yet to find the amount of light for the whole sum expended nearly the same.

The best relative proportions of oil and oxygen for the Good Lamp are about One pint of the former for Ten cubical feet of the latter. It is not the proportions only, but the relation, also, in which the two are placed by position in the flame, which are essential to the production of a good result. Thus in the table at page 61, the results with the No. 5 lamp on the 24th and 31st August were with oil and oxygen nearly in that ratio; but the price of light equal to one Argand burner was 15.6 pence instead of 10 pence, as by the Good Lamp. (page 65.)

When the lamp is burning with smoke the fuel may be considered as in excess. If the supply of oxygen be then increased the light rises much in intensity, the smoke diminishes, the flame is shortened and if the supply of oxygen be still increased, the smoke altogether disappears, and the oxygen, after that, may be considered as in excess. It is desirable that the fuel should not be in excess since then it is going to waste, and the whole light, instead of being increased at such times, with increase of fuel, is occasionally actually diminished by it. The results on the 8th and 18th of October in the table at page 65 illustrate this point. In practice I expect it will be desirable to adjust the constituent flames so that they shall yield a very small quantity of smoke. If they are in such a state as to yield none, it may happen that oxygen will be passing in excess through the flame without the knowledge of the light house keeper, and so be running to waste.

The purer and better the oxygen the better and more economical is it’s effect in the lamp. All depends upon having the most intense heat in the focus of action, and that depends upon the purity of the gas. Air is only diluted oxygen; but being diluted, it cannot, either in the common Argand or in the French lamp, produce the intensity which is obtained in the Good Gurney lamp; and when, as in the French lamp, intense focal combustion is sought to be obtained by it’s use, that which is attained is accompanied by a wasteful expenditure of fuel.

Many experiments tend to shew, as may be seen by the tables pages 61, 65, that the higher the illuminating powers of the oxygen lamps are raised the cheaper is the light for equal quantities of the latter. Thus, the lamp No. 5, when giving light equal to nearly 32 Argands, did it at the cost of 14.5 pence per Argand; whilst when giving the light of 21.23 and 19.8 Argands, the cost respectively was 15 and 16.3 pence. So also the Good Lamp when giving the higher proportion of light, did it at the cheapest rate.

The form of light given by the Good Gurney Lamp seems to me to be a very advantageous one. It is four inches in horizontal width and only 1¼ to 1½ inches high. Hence when placed in the focus of a lens the horizontal divergence of the refracted beam will be several times greater than the vertical divergence and in that respect very fit for practical purposes.

When a lamp flame 3½ inches in diameter is placed in the focus of a lens, as in the French dioptric lighthouse arrangements, much of the light in that part of the flame furthest from the lens, is prevented from falling on the lower part of the glass, being intercepted by the intervening parts of the cotton and burner. This cannot be helped in such arrangements as the French lamp or Gurney’s Annular lamps; where the oil rising to a horizontal level, is in communication with all parts of the same large ring. But in the Good Gurney lamp the constituent burners may be placed on two or more levels; and in cases where the lighthouses are not required to illuminate all round the horizon, those constituent burners on the land side of the lamp, may be raised above the others, and so all be made to shed their beams over the whole of the lens.

The light of the lamp is very pure and white in colour. As yet I believe only a red colour is given by glasses to the light from reflectors; but it is a question whether with this light, in consequence of it’s whiteness, blue or green tints may not be obtained with advantage, either in catoptric or dioptric arrangements.

It seems to me very probable that the lamps already in use burn with very different degrees of lustre at different times; and with the French lamp, as I have seen it burn, this result appears to be inevitable. The same conclusion is justified by the variable proportions of oil which are burnt in the Menai and Start Point lamps; the quantities at both places differing from those which I or Lieut. Drummond have obtained for the same kind of lamp. In Gurney’s Good Lamp I expect that considerable regularity in this respect may be ensured. At the commencement I would recommend that the consumption of oxygen in a given time be fixed; as for instance at 8 or 9 or 10 cubical feet per hour. This can easily be done by a proper stopcock or valve with an index lever and graduated scale attached; and the fit supply of oil can be adjusted with facility by the arrangement of the wicks, care being taken to burn it thoroughly according to the principles and statement already made. (pages 71, 72.). In this manner a scale of combustion can be given to the light-house keeper, and he will be enabled to sustain with regularity a larger or a smaller light with the same lamp according to the instructions given to him by the Board.

For the purpose of checking the light-house keeper; and also for obtaining accurate data respecting the performance of the lamp, the production and price of oxygen, &c., &c., it will be proper to have a gas-meter between the retort & the gas holder, and another between the gas holder and the light; both being locked up, so that the keeper shall be able to observe, but not to alter, them.

Supposing a lamp equal in power to 20 Argands, were set in a Lighthouse, it would require about 64 cubic feet of oxygen for 12 hours, which would require for it’s production 87.36lb of Warwickshire Manganese; and taking 12 hours as the mean time of burning for the whole year, the annual consumption of Manganese would be about 15 Tons.

<-><-><->

Before concluding this Report, I am called upon to state, that since I have entered upon the practical investigation of the merits of the Gurney lamps, Sir David Brewster has informed me, that as early as the Year 1831 he suggested to the Engineer of the Scottish Lighthouses the supplying of oxygen in place of atmospheric air to the lighthouse lamps; and that in April 1833 he observed in No. 115 of the Edinburgh Review that “the lens lamp might be supplied with oxygen gas in place of atmospheric air.”⁵ This is of course an announcement of the principle of the Gurney lamp, and long before that time chemists had been occasionally in the habit of supplying oil lamps with oxygen to illustrate the intense light and heat produced. Having said this I must also add that as far as I know Mr. Gurney is the first person who has practically applied the principle to lighthouse lamps, and worked it through all it’s difficulties to a successful conclusion. What these difficulties were are sufficiently obvious from the various reports I have had the honor to submit respecting the lamps, successively placed before me during the last three Years⁶.

I have the honor to be | My Lords and Gentlemen | Your obedient faithful Servant | (signed) M. Faraday

Royal Institution | 29th October 1838.