800 THE COLLIERY GUARDIAN April 28, 1916. the volts across the lamp (2’1 volt and 0’9 amperes) were 1*2 and total amperes in circuit equal 0-68. Without iron wire in parallel with platinum filament the volts across lamp were 1'0, and amperes in circuit, 0-62. (2) Using same cell, but T3K in. platinum wire 0'0015 in. diameter in series with lamp. Volts across lamp were 0'8, and amperes in circuit, 0*57. The platinum fila- ment was glowing fairly brightly and lamp just also glowing. Now inserting 8 ft. 9 in. of 29 S.W.G. steel wire in parallel with the platinum wire the volts across the lamp rose to 1'1, and amperes in circuit, 0’65. The lamp just slightly brighter, but still only glowing, and platinum filament just lowest red visible. (3) Reducing the length of steel wire to 5ft. 10in., the volts across lamp were 1'2, and amperes in circuit, 0'68. No light in platinum filament, and not much on- lamp, on passing gas (coal gas) past filament it incandesced and became bright red, and then votes across lamp equal 1’2, and amperes in circuit, 0'67. The greatest drop that we could get with gas was a variation from 0'68 amperes to 0'65 amperes. (4) Using a 4 volt cell of large capacity, and with 74-in. length of 0*0023 in. diameter platinum wire in series with the lamp, the volts across lamp were 2'7, and 1*2 amperes in circuit with 5 ft. 10 in. of 29 S.W.G. steel wire in parallel with filament and all resistance in, the lamp burnt brightly, but platinum filament not red; now cutting out all the steel wire, but rest same as before, the volts across lamp fell to 2’5, and amperes in circuit equal 1’15. Without the steel wire just as above the drop in amperes on passing gas (coal) till it inflamed, was from 1’15 to 1'10, there was no visible effect on the light of the lamp. (5) Using about -3 in. length of 0'0023 in. diameter platinum wire on same 4 volt cell, the volts across lamp were 2'1, and amperes in circuit, 0'9. Now on passing coal gas until it inflamed past platinum filament the amperes dropped to 0*89, and volts across lamp practically the same. No variation visible in the light of the lamp on passing the gas till it inflamed. (6) Much greater variations caused by currents of air as blowing directly on the platinum wire, thus by blow- ing directly on the platinum filament the maximum variation in volts across the lamp was 2'1 hot to 2'3 cold, and amperes varied from 0'90 to 0'94. Volts across cell, 4'1. (7) Using 4 in. length of 0'0023 in. diameter platinum wire in series with lamp volte across cell equal 4’1, amperes in circuit 0'89, and volts across lamp 2’1. Placing steel wire in parallel with filament, amperes rose to 0'96, and volts across lamp were 2'5. It then became necessary to increase length of platinum filament to lin. to get volts across lamp down again to 2'1, and amperes in circuit 0*88, and then with this condition (all. resistance also in) the drop was only to 0'87 amperes when gas (coal) was passed till it inflamed. (8) An endeavour was made to see greatest variation possible in current passed by -Jin. length of 0'0023 in. diameter platinum wire. So the platinum wire was wired directly in series with the resistance, and then on the cell. The lamp was not in circuit. Volts across platinite filament 1-25, and ampere 0-99, and on blow- ing till filament changed from bright red to cold amperes rose to 1-09. In the above all resistance was in. With one cut out volts across filament = 1-5, and amperes 1-08, and on blowing amperes rose to 1-29. With two cut out volts 1-6, and amperes 1-14, and on blowing amperes rose to 1-38. With all cut out volts 1-75, and amperes 1-16, and on blowing rose to 1-43. The increase in amperes on cooling therefore from bright red to cold varied from 0-1 ampere, 0-11 ampere, 0-14 ampere, to 0-27 ampere in last case. (9) With | in. length of 0-0015 in. diameter platinum wire in series with lamp, and placing two strands of 4| in. steel wire 29 S.W.G. in parallel with the platinum wire, and using a large 2-volt cell, the amperes in steel wire were 0-60, and volts across lamp 1’3. Lamp just glowing, and platinum filament not red.. All resistance in. Cutting out all the resistance, the amperes rose to 0-75, and volts to 1-6, the light a good deal brighter, but still of no use as an illuminant, no redness in platinum filament, and no effect in either of the above cases on passing gas (coal). (10) An endeavour to show the least amount of current necessary to heat —r in. length of 0-0015 in. diameter platinum wire in order to incandesce in coal gas. The wire was wired directly on to cell in series with a large steel wire resistance :— 0’17 amps no effect. 0’22 „ „ „ 0-26 ,, „ 0-275 „ „ „ 0’283 „ „ „ 0’29 „ „ „ 0'295 showed effect with gas. Repeated 0'29 ,, ,, ,, >, 0'28 ,, no effect. 0'29 dropped to 0'265 in passing gas. 0'33 „ „ 0'29 „ 0'35 „ „ 0'31 ,, 0-37 „ „ 0'33 „ So minimum current is 0-29 ampere for y3^ in. length of 0-0015 in. diameter platinum wire. Experiments with Methane. (11) This test was carried out to show minimum current required to react with methane, using about in. platinum wire of 0-0015 in. diameter. The cell used was a small 2-volt cell of about 9J ampere hour capacity. The gas was allowed to enter before each reading of the ammeter and voltmeter, but no special care was taken to clear it out between each reading. There was no apparent action till the volts rose to 1-95, and ampere passing through wire was 0-505, then the wire was observed to get brighter when the gas was passed into apparatus, and there was also at this time the first indication of variations in the amperes. The variation was caused by using an iron wire resistance, and gradually cutting out same. Amps. 0'285 Volts across plat. wire. 0'65 Observation. No effect 0'3 0'8 0'345 0'97 0'4 1'25 0'435 1'45 0'455 1'55 0'475 1'65 7, (At this point filament first showed least redness) 0'505 0'505a . .. ■ 1'95) 1'95 J '" Filament got much brighter and amps, dropped momentarily 0'475 to about 0’42, and then remained steady at (12, 13 and 14) Experiments with a 10 ampere hour capacity 4-volt cell, using copper wire coils in series and in parallel, and giving results. Copper increases about 0-238 per cent, in resistance per deg. Fahr. These experiments were carried out with a view primarily of reducing the resistance in one case, and of increasing it in the other, and of accentuating the reduction in current reaching the lamp, when wire becomes further heated in gas, and thus increases the resistance. The most suitable wire for this cell, viz., 4-volt, is the 0-0023, by which we were able to maintain the usual light of the bulb, and at the same time incandesce the platinum filament. Volts across cell showed 3-85, volts across lamp, 2-0, ampere in circuit 0-9, using about -Jin. length of the platinum wire. The effect of using the copper wire coil in parallel with the platinum filament while the filament and lamp are in series, is to take part of the current away from the platinum wire, depending on the amount of copper wire used. It increases the amount of current used. Thus, Jin. of 0-0023 in. diameter platinum wire in series with resistance and lamp gave 2-15 volts across lamp, and 0-95 ampere in circuit, and inserting 16 in. of 0-015 in. diameter copper wire (about 28 S.W.G.) in parallel with the platinum filament, the volts across the lamp rose to 2-55, and amperes in circuit were 1-05. 6- Fig 1.—Wood and Glass Box. With platinum wire, lamp, copper coil (16 in. long and 0-015 in. diameter) and resistance. All in series. The volts across the lamp were 2-0 and ampere in circuit was 0-90;, light was bright, and platinum wire was red. On adding a further 19 in. of same copper wire, making 35 in. in all, the volts across lamp fell to 1-75, and ampere in circuit was 0-85; the light of the lamp had fallen considerably, and the platinum wire showed just visibly red. Effect of Reduction in Voltage on Filament Lamps. The following figures show what an effect a drop in voltage has on the candle-power of a small electric lamp bulb Volts. Candle-power. Original candle- power retained. Per cent. 2T 1'27 100 2’0 1'00 79 1’9 0'85 67 1’8 . 0'67 52'7 1’7 0'555 43'7 1’6 0'416 32'8 1’5 0'296 23’3 1'4 0'208 16'8 An ordinary 4 volt cell, weighing about 41b., will run the 0'0023 in. diameter wire (| in.) in series for over 8 hours, and give a light of over 1 candle-power. A 15 ampere hour Alklum cell will run the 0'0015 in. diameter wire (T%in.) for 10 hours, the wire being in parallel, and give a light of 2 candle-power for 10 hours. The voltage of this cell is 2'5, and its weight about 4J lb. An ordinary mining lamp 2 volt cell weighing about 3 1b. would run the 0'0015 in. wire (T3^in.) and a 0-75 ampere 2 volt lamp in parallel for 10 hours, and give about 1 candle-power. Probably a suitable “ Fors ” 4 volt accumulator could be made suitable for a mining lamp—with regard to weight—which would give 1 candle-power with a filament lamp for 10 hours with the lamp and 0'0023in. diameter wire in series. Grimsby Coal Exports.—The following is the official return of the quantities of coal exported at Grimsby during the week ended April 20 :—To Civita Vecchia, 2,125 tons ; Dieppe, 554; Esbjerg, 344; Gothenburg, 1,684; Haugesund, 109; Savona, 2,563; and Treport, 623—a total of 8,002 tons, com- pared with 12,719 tons in the corresponding week of last year. The Neglect of Science.—A meeting will be lield, at the invitation of the Committee on Neglect of Science, on Wednesday, May 3, in the rooms of the Linnean Society, Burlington House, Piccadilly, London. Lord Rayleigh, O.M., past-president of the Royal Society, and Chancellor of the University of Cambridge, will take the chair at 3 p.m. The resolutions to be submitted recommend that natural- sciences should be made an integral part of the educational course in all the great schools, and should form part of the entrance examinations of the universities; that the Govern- ment should assign capital value to these sciences in the examinations for Home and Indian Civil Service; and that a committee should be appointed to bring the matter urgently before the Government. Those desiring admission to the meeting should apply to the hon. secretary, 28, Victoria-street, Westminster. LIMITS OF INFLAMMABILITY OF MIXTURES OF METHANE AND AIR* By G. A. Burrell and G. G. Oberfell. Experiments have been made at thef Pittsburg experi- ment station of the United States Bureau of Mines on self-propagation of flame in mixtures of methane and air. The literature on the inflammability limits of methane- air mixtures is voluminous. Below are shown limits obtained by different investigators :— Inflammability Limits of Methane-air Mixtures. Percentage of methane. Investigator. f Lower limit. Upper limit. Davy 6’2 to 6’7 14’3 Clowes 5'0 13’0 6’0 11'0 Eitner 6’0 to 6'2 ... 12'7 to 12'9 6'3 — 5’5 — Teclu 3’2 to 3'67 — Perm an 2'5 — Burgess and Wheeler 5’6 — Clement 5'8 12’8 5'5 11'7 Leprince-Ringuet 5’76 — 5’56 — 5:20 — Coward and Brinsley 5’30 — 5'90 — 5'30 — 5’40 — 6’20 — 6*50 — Of these, the extremely low values obtained by Teclu and Perman have been the subject of criticism, and are regarded as erroneous. It is evident that the limits of inflammability of any given mixture of methane and air cannot be definitely fixed at one value, so as to cover all conditions of experi- ment. That the diameter of the tubes may be so small as to cool the flame and hinder combustion has been shown by Mallard and Le Ch atelier, -working with methane and air, and by Le Chatelier, working with acetylene and air and other mixtures. The diameter of the tube necessary to avoid cooling by the walls, and consequent retardation of the flame was found to be greater the slower the speed of the flame. For the most slowly moving flames in mixtures- of methane and air a tube at least 5 cm. in diameter is necessary. The speed of travel of flame in a tube 9 cm. in diameter is slightly greater than that in a tube 5 cm. in diameter. Le Chatelier found that the maximum speed in methane- air mixtures was not obtained with a mixture contain- ing 9-47 per cent, of methane—that is, the quantity of methane required for complete combustion—but with a mixture containing 12 per cent, of methane. Wheeler repeated Le Chatelier’s experiments, but did not con- firm the 12 per cent, value. He found that there is practically no difference between the speeds attained in mixtures containing 9-45 to 10-55 per cent, of methane. That initial pressures und temperatures may be rather high without appreciably affecting the limits is shown by recent work at the Bureau of Mines laboratories. Determining the Low Limit of Inflammability. Methane for the authors’ experiments was obtained from a natural gas well near Pittsburg, Pennsylvania. Steel cylinders were filled at the pressure of the well, 1901b. per sq. in., and brought to the laboratory. Analysis of the gas showed it to contain 98-8 per cent, of methane, a trace (0-03 per cent.) of carbon dioxide, and nitrogen. In experiments with a wood and glass box and with an iron pipe, the gases were partly saturated with mois- ture. In experiments with smaller glass containers, they were completely saturated at the temperature of the laboratory. A wood and glass box (fig. 1), 5| ft. (1-75 m.) high and 1ft. (30-5 cm.) square, and having a capacity of 53 cu. ft. (163 litres), was used to confine some of the mixtures. A fan placed at the extreme bottom of the box served to mix the gases prior to ignition, which was caused by drawing apart two copper wires, through which an electric current of 7 amperes at 220 volts was flowing. The wires were placed at the bottom of the box, just above the fan. Samples of gas taken both at the bottom, and at the top of the box showed that the fan mixed the gases thoroughly. An aperture at the top of the box was covered with a tight paper diaphragm. Upon ignition of the gas, the diaphragm broke, and gave a vent for the burned gases. Results with Ignition from the Bottom. In experiments in which the large box was used, igni- tion being from the bottom upward, the lower limit was about 4-9 per cent, of methane. This value is consider- ably lower than that of Coward and Brinsley, and is contrary to the statement of Burgess and Wheeler that the lower limit of upward propagation of flame contains not less than 5-4 per cent, of methane. * From Technical Paper 119, U.S.A. Bureau of Mines.