Ignition of Go,ses by Hot Wires. 621 



at say 1300° 0. absolute, which is about the observed bright- 

 ness, is, taking the vessel at 27° C, given by 



H c is here 29,000 calories, d = '01 centimetre. 

 From which 



m = 1-255. 10- 4 + 3-82 /lO" 6 - M8, 10" 7 - 5*23. 10" 5 



= '0000524 gram per square centimetre per second 



l-b\3 - , 



= Yffi gram per linear centimetre per second. 



The explosion vessel filled with a 30 per cent, mixture 

 contains 15 c. cm. of hydrogen or '00135 gram. To burn this 

 by a 3 cm. length of wire from 30 per cent, to 10 per cent, 

 of mixture takes '00089/4'89 . 10 <; seconds, that is 182 seconds. 

 The surface combustion glow should therefore last three 

 minutes to complete blackness again, and ihis is much the 

 time observed though the final extinction is not well 

 marked. 



In certain cases of rich mixtures the electric heat may be 

 cut off and the wire continues to glow by surface combustion 

 alone. The duration of the glow is then less than when the 

 wire carries a current. 



9. Temperature of Wires at Ignition. 



The least igniting current is that which starts gaseous 

 combination at such a rate that explosion follows. In every 

 case there is surface combustion which may proceed at a 

 steady rate without giving rise to explosion though the wire 

 is glowing brightly. The phenomena before the wire glows 

 in an explosive mixture are extremely interesting. The wire 

 is in almost every case darker in appearance than at the same 

 electric current in air alone. In hydrogen and benzene par- 

 ticularly, the effect is marked. I'£ before explosion the 2 mm. 

 tap at the top of the explosion vessel is opened so that air 

 can be drawn in after explosion by contraction of the pro- 

 ducts, the wire glows more brightly, and continuously, after 

 explosion than before, the current remaining constant. 



A sensitive voltmeter connected across the filament shows 

 that as the current is increased the resistance begins to rise, 

 slowly at first and then with a rush, the current meanwhile 

 changing very little, from a point at which the wire is 

 apparently cold. At ignition the wire appears incandescent, 

 yet the readings of the voltmeter indicate a temperature 



