258 THE COLLIERY GUARDIAN. August 11, 1916. It will also be seen that the temperature of the sur- rounding mixture of gas and air caused a variation. With a 7 per cent, mixture, there was a rise in the voltage to 2-02 from 2’0 volts (see experiment 11). With a 25 per cent, mixture there was also a rise from 1’8 to 1-82 volts. This variation was caused by the gas issuing from the cylinder being extremely cold, as the result of sudden expansion, and of the gas mixture being in contact with cold water in the mixing apparatus, so that it did not have much chance of attaining the tem- perature of the laboratory. It will be noted in experi- ments 21 and 22 that the brightness of the wire was considerably increased, and that it was visible in day- light (even with bright sunlight streaming in through the windows of the laboratory). The variation was only 0-06 volt 0-03 amp. in experiment 21, and 0’06 volt 0’035 amp. in experiment 22. There was also a varia- tion of 0’02 volt and 0’01 amp. with the 7 per cent, mixture. I • 5 Amp. shunt Reg Resistance Reg.Resistance 0"3 Volts Platinum Wire -O-----O’ Arrangement of Apparatus in Testing. Ex peri-1----- ment’ Diam. In. 1 ... 0’0023 2 ... 0’0023 3 ... 0’0023 4 ... 0 0023 5 ... 0’0023 Platinum wire used. ____________________________________________________________________________________________________________________________________________________ GAS DETECTOR FOR MINERS’ ELECTRIC SAFETY LAMPS. By T. J. Thomas. After the publication of the previous articles* relating to experiments with platinum wires in methane and coal gas, the author drew the attention of one of the most prominent coal owners in South Wales to the question, and it was considered advisable to carry out a further series of experiments, which .were conducted at the ’South Wales and Monmouthshire School of Mines, Treforest, by kind permission of the principal, Prof. George Knox, F.G.S., M.I.M.E. The results of these experiments show that no catalytic action takes place with platinum wires heated, to a bright cherry red heat in a 20 per cent, mixture of methane and air, but that catalytic action does take place with 25 per cent, mixture of methane and air, and that this catalytic action is so violent with 33J per cent, mixtures of methane and air, that the platinum wire is immediately fused. This refers to catalytic action which would give a variation' on a sensitive voltmeter and ammeter, whilst possibly a variation would be obtained on a galvanometer. There was a variation (a drop in voltage) on the volt- and ammeter with a 7 per cent, mixture, but the variation in this case would be caused by the combustion of the gas, and the consequent heat ______________ would raise the temperature of the wire and so increase its resistance to the passage of the current. One point to be noticed in these experiments is that a length of about fin. of 0-0023 in. diameter platinum wire would act as a suitable fuse on a 4-volt cell using a 0-85 amp. lamp, in the event of short circuit in the cell. The wire is near fusion when a current of 1-06 amp. at 3 volts is passed across the lamp (see experiment 19), whilst in experiment 23 it fused at 2-8 volts (across lamp), and with 1-02 amp.; but in this case there was 25 per cent, of methane in contact with the platinum wire. It will also be seen from experiments 20 and 21 that the wire could be made to fuse, and thereby cut out the light before a mixture of gas and air capable of causing asphyxiation would be reached. The wire fused with a mixture of 33J per cent, of methane at a point which would represent the lowest voltage on a mining lamp cell at end of discharges—i.e., a 4-volt cell under the above conditions. In a mixture of about 28-30 per cent, the lamp would probably behave as in experiment 12. There was a drop in voltage in this case of 0-64 volt, and a fall of 0-26 amp., from 2-08 volts 0-09 amp. to 1-44 volts 0*74 amp., when .the wire fused. This would represent a drop in candle power on a fila- ment lamp (miners’ lamp) from about 1-2 to about 0-25 to candle power, or a drop of nearly 0-8 in the light, a change which could not fail to be noticed. Assuming that the firedamp was almost entirely com- posed of CH4, a mixture with 33J per cent, of methane and air would give : 14 parts oxygen, 52J nitrogen, and 33| methane; and a 25 per cent, mixture of methane and air would give an atmosphere consisting of 15’75 oxygen, 59’25 nitrogen, and 25 parts methane. Methane has the same physiological effect as nitrogen, and in this respect merely acts as a diluent of the oxygen of the atmosphere. Table I. __________ Combustible sub- Percentage composi- cen^ of o2 and stance burnt. tl(^’ N^.f______ _________ ^O. N. CO^O. N.~^ Colza and paraffin... 16’4 ... 80’5 ... 3T ... 16’4 ... 83’6 Candle ......... 15’7 ... 81’1 ... 3’2 ... 16’4 ... 83’6 *Of the residual atmosphere in which the flame is ex- tinguished. fin which flame is extinguished when introduced. In the tablef published by Prof. Clowes, D.Sc., of the percentage composition of residual and of arti- ficial atmospheres, which extinguish flame, the following results are shown in Table I. Accord- ing to the statements published by Dr. J. Haldane (Proc. Roy Soc., December 6, 1894), an atmosphere of the average composition of expired air or that left by the combustion of candles and lamps-can be breathed by most people without producing any dis- tinctly noticeable effect. With some people the increased proportion of carbon dioxide would cause the inhalation to be rather deeper than usual, but in no case would injury to health result from breathing such an atmosphere. The percentage of methane in the gas (from Cyrnmer Colliery) used in the present experi- ments was 97-5 per cent., the remainder (the greater part of it) being nitrogen and the other portion CO2, the firedamp in this respect being similar to the gas from Dunraven, Vochriw, and Llwynypia, South Wales, and shown in Table II. Table II.—Analysis of Firedamp from several South Table IV.—Action of Heat upon Ethane (giving only amount of hydrogen.) (V. B. Lewes ) Percentage of ethane in original gas ......... 96’38 ... 96’38 ... 96’38 ... 96’38 Temperature of gas in the decomposing tube 900°C. ...l,000°C....l,200°C. .l,500°C. Hydrogen ............. 40’64 ... 46’78 ... 57’45 ... 73’35 Table V. Experiment to ascertain length of 0’0023 in. diameter of platinum wire most suitable for running in series with a (tungsten) filament 2 volts 0’85 amperes lamp. Apparatus Used.—Two chloride accumulator cells, each having a capacity of (at eight hours discharge) 2u0 ampere hours at 2 volts. These cells were connected in series, giving an E.M.F. of 4 volts. The 2 volt 0’85 ampere filament lamp was connected in series with the platinum wire. A combined precision voltmeter and ammeter, for simultaneous current and pressure measurements (type Midr. 73) was used for the electrical measurements. Two large regulating resistances were used for regulating the current to the desired quantity. Using the 1’5 ampere scale of the precision ammeter (150 divisions) a reading to 0’01 amperes could be taken. Using the 3-volt scale of this instrument (150 divisions), a reading to 0’02 volts could be taken. Fahrenheit thermometer used 10 to 400 degs. Fahr. Note.—Large cells were used to prevent drop in volts due to low capacity and consequent comparatively high resistance of cells. Temperatures Electrical (Eahr.) measurements Length. Lab.J 4ir nea: wire. Volts. Amps. In7 Mm? ... 13 Degs. ... 70 . Degs. Degs. .. — ... 2’2 ... 0’93 30 "8 4 • .. 13 ... 70 . — .. — ... 1’8 ... 0’84 25. 6 4 • .. 10 ... 70 . " 166 .. 96 . .. 2’125 ... 0’93 25. fl 4 • ■ .. 10 ... 71 . .. 158 . .. 87 .. .. 2’0 ... 0’9 25 64 •• .. 10 ... 72 . .. 147 . .. 75 .. .. 1’8 ... 0’85 Remarks. In Experiment No. 1, all resistance out direct across cells. Platinum wire showed red (in daylight) In Experiment No. 2, platinum wire just glowing red in daylight. In Experiments Nos. 3, 4 a? d 5, thermometer -?2in. away from wire. Current allowed to pass through wire for 10 minutes and thermometer observed to steady. Wire screened from draughts. Wire glowing very bright red. Experiment 5 represents a 10 per cent, drop in volts. Temperature of platinum wire considered correct for re- acting with methane, and voltage and current for filament lamp (approx.) correct, being only 0’05 ampere above rating. Table VI. Experiment with 0 0023 in. diameter platinum wire in methane. (The methane obtained from a cylinder of the compressed gas obtained from Cyrnmer Colliery, Forth.) The pressure of the gas in the cylinder being about 225 lb. per square inch (15 atmospheres). Apparatus generally as in Experiment 1, 2, 3, 4 and 5, but with the addition of a bell jar 4| in. diameter and 8 in in height. The tubula- s being closed with a cork, the bottom being left open. A small quantity of methane was allowed to pass into bell from a tube connection to the gas cylinder, and the lamp and platinum wire arrangement inserted at a point about 7 in. from the bottom opening of the bell. Conditions: Daylight, bright sunshine streaming in through the windows. Note the temperature of the air increasing in the laboratory. Platinum wire used. Temp, in Experi- (---------A-----------> Labora- ment Diam. Length Length tory. In. In. Mm. Degs F. 6 ... 0’0023 ... || ... 10 ... 73 . 7 ... 0’0023 ... || ... 10 ... 73 . Remarks. Electrical measurements. Volts. Amperes. 2’0 ... 0’88 2’1 ... 0’89 In Experiment No. 6 same wire was used as in Experiment No. 3. Admitted methane (small quantity) to a point near top of bell. Platinum wire almost immediately snapped (through overstrain by contraction in cold gas from cylinder), or fused, thereby breaking circuit. In Experiment No. 7 wire (platinum) renewed with wire fresh from bobbin. Admitted im thane as before (Experi- ment 6), but in smaller quantity. Wire glowt d less brightly at first, owing to reduction in temperature due toadmission of methane from cylinder. Result.—Wire gradually increased in brightness. Wire increased in temperature from bright red to orange bordering on whiteness, as the methane mixed with the air in the bell. The voltage across lamp dropped from 2’1 to 1’66, a drop of 0’44. The drop in candle-power was not distinctly observed on the lamp, but would be from 1’27 candle-power at 2’1 volts to 0’555 at 1’7 volts with a 2-volt lamp. At 1’66 volts the platinum wire behaved in a most extraordinary way. First the lamp filament went completely dark, and half of the platinum wire appeared dark, the other half being nearly white, and then completely dark. Electrical instrument showed zero, and on examination platinum wire was found broken, either snapped or fused. Table VII. Experiment with 0’0023 in. diameter platinum wire in series with a 2-volt 0’85-rampere lamp (tungsten) in methane and air. Method : Passed a small amount of gas direct from cylinder into bell, and allowed about 5 minutes to mix with the air. Bell, 8 in. high, (internal) diameter 4£ in.; open at bottom. Wire about half-way up in the bell. Apparatus connected up as in Experiments Nos. 1, 2, 3, 4, 5, 6, 7. Wales Mines.* Mine. CH4. N. O. CO2. Various. Dunraven... 97’65 ... 1’85 ... 0’0 ... 0’05 ... — Vochriw ... 97’37 ... 2’21 ... 0’0 .. 0’42 ... — Fforchaman... 95’05 ... 0’69 ... 0’0 ... 4’26 ... — Llwynypia ... 94’78 ... 3’60 ... 0’0 . 0’72 ... 0’90 C2H6. *From “'Coal, Mine Gases, and Ventilation?'’ (J. W. Thomas.) It will be observed that the catalytic action increases in proportion to the increase in temperature of the platinum wire (see experiments 14, 15, 16, 21, 22, and 23), and that the temperature of the wire can be varied by lengthening or shortening it, or by varying the current passed. This is shown in experiments 1, 2, 3, 4, and 5, and also very clearly in experiments 11 and 19. The variation due to combustion of the gas in mixtures of methane and air will of course fluctuate with the percentage from 1 to 9-38 per cent, of methane, the greatest variation being at 9-38 per cent. The rate of combustion can be increased by raising the temperature of the wire, and the variation in the instrument readings will be increased proportionately. * Colliery Guardian, Jan. 28, IQieTpTTT^April 28^1916, p. 799. f Proc. Roy. Soc. Ixvi-i, 573. The experiments were carried out with the platinum wires at cherry red and bright cherry red heats—the former when the voltage (across lamp) was 1’8, and the latter when the voltage was 2-0 and 2’15 volts. These temperatures are approximately 910 and 1,000 degs. Cent., so that this is the temperature at which the smallest amount of hydrogen is formed during the action of heat upon methane. The intensity of the catalytic action upon the platinum depends, of course, upon the amount of hydrogen present, and it therefore follows that the catalytic action will increase as the tempera- ture of the wire is increased, so that by raising the temperature of the platinum to a white heat (1,300 degs. Cent.) the amount of free hydrogen liberated would be about seven times as great, and at a dazzling bluish white (1,500 degs. Cent.) the amount of hydrogen would be over 50 times as great as at a temperature of 1,000 degs. Cent. The author considers that a temperature of about 1,400 degs. Cent, will be about the best at which to keep the platinum wires heated, since the fusing point of platinum being 1,775 degs. Cent., this would allow for an increase in temperature of 375 degs. Cent, before the wire would fuse. It will be observed in experiments 18 and 19 that the wire was not in imme- diate danger of fusion when the voltage (across lamp) was 3’0 volts, and the amperes in circuit 1’06. The wire was then dazzling white. It may be mentioned incidentally that the working temperature of the tungsten filaments used in miners’ lamps is approxi- mately 1,400degs. Cent., so that the platinum filament and tungsten would be at the same "temperature. It may be found advisable to work the platinum at this temperature, in which case, by contrasting the two fila- ments, very small variations of temperature could be noted in the platinum, and therefore, small percentages of methane detected. It must not be forgotten, how- ever, that the platinum is in direct contact with the air, whereas the tungsten filament is enclosed in a sealed globe, so that a rise of, say, 30 degs. Cent, in tempera- ture may possibly give an indication similar to a small percentage of methane, say, 1 or 14 per cent. This fact appears to be lost sight of in certain electric gas detect- ing devices. Table III. (action of heat on flowing methane) shows that a trace of carbon is deposited at 1,300degs. Cent., and 0’015 per cent, of carbon at 1,500 degs. Cent. It is possible that the deposition of this carbon on the platinum wire causes a variation and the change in the electrical resistance of the wire in such instruments as the Liveing apparatus, and thus makes the readings inaccurate after the platinum has been used for a time for testing and in contact with methane. The author has found that the wire has a tarnished or smoky appear- ance after being used in contact with methane, though it should have remained bright, as it doos not oxidise in air until brought to a state of fusion. The author pro- poses to change the wire in his detector when it shows a variation in resistance. This can be effected very economically. Firedamp sometimes contains a small percentage of ethane (and occasionally the quantity is large, as in silver gas), as, for example, in the sample from the Glamorgan Colliery, Llwynypia. Ethane will probably show a different effect on the platinum wire. The catalytic action will probably be much greater, owing to the large amount of hydrogen produced bv heat. (See Table IV.). Table III.—The Action of Heat upon Flowing Methane. (V. B. Lewes.) No. Platinum wire Labora- vinn4r;nni Of • USX measurements. experi- ( A tempera- > Remarks. ment. Diam. Length, ture 1 In. In. Degs. F. Volts. Amps. 8 ... 0’0023 ... | ... 67 ... 2’1 . .. 0’9 .. Res. out a ... 0 0023 ... il ... 67 ... 1’8 ., .. 0’84 .. .. Res. in 10 % drop. B 1’8 . .. 0’84 9 ... 0’0023 ... 67 ... 2T . .. 0’9 . .. Res. out 10 ... 0’0023 ... 11 ... 68 ... 21 . .. 0’89 . Percentage of methane in the original gas ... 99’2 .. 99’2 .. . 99’2 ... . 99’2 Temp, of gas in the decomposing tube 1,000°C.. ,.l,200°C... l,300°C...l,500°C. Analysis of the gas after heating— Unsaturated hydro- carbons trace . .. 0’07 .. . 0’39 .. . 1’20 Containing acetylene.. trace . .. 0 07 . .. 0’39 .. . 0’963 Saturated hydroerbns 97’65 . .. 90’00 . .. 88’52 .. . 19’22 Hydrogen 1’55 . .. 8’53 .. .. 10*37 .. . 78’66 Carbon deposited and oil formed in grammes per 100 c.c. of gas— Carbon 0’0 .. 0’0 ,. trace .. . 0’015 Oil 0’0023. .. 0C025. .. 0’0005.. . 0’0 Remarks. In Experiment No. 8 (observed in daylight), platinum wire very bright red, filament lamp brilliant; a, platinum wire fairly bright red, filament bmp brilliant; b, passed small amount of methane from cylinder into bell, and allowed about 5 minutes to mix, and placed over wire. Drop to 1’7 volts 0’82 amperes, then rise to 1’73 volts 0’825 amperes, then rise to 1’8 volts 0’84 ampere. Total (highest) variation, OTO volt 0’02 ampere. In Experiment 9, repeated generally as above, platinum wire very bright red. Placed bell over wire. Drop to 2’04 volts, drop to 1’98 volts 0’87 ampere, drop to 1’96 volts ; lowest, 1’94 volts 0’86 ampere. Volts and amperes began to rise here. Total (lowest) variation, 0’16 volt 0’04 ampere.