678 EEPOET— 1892. 



experiment is to allow one or two drops of a solution of SOo in alcohol or ether to 

 fall on a little powdered potassium chlorate ; white fumes are given off which are 

 nearly always followed by a sharp explosion. 



8A1URDAY, AUGUST 6. 

 The Section did not meet. 



3I0NDAY, AUGUST 8. 

 The following Papers were read : — 



1. On the Application of a Hydrogen Flame in an Ordinary Safety-lamp to 

 the Detection and Measurement of Inflammahle Gas or Vafour. By 

 Professor Frank Clowes, D.Sc, University College, Nottingham. 



The appearance of a ' cap ' over the flame in the safety-lamp has long been 

 used by the coal-miner for detecting ' firedamp' in the air, and for roughly measur- 

 ing its amount. The ordinary oil-flame does not with certainty detect the presence 

 of less than 3 per cent, of firedamp. The alcohol flame adopted by Pieler detects 

 0'25 per cent, readily ; but since this flame gives no light the Pieler lamp can be 

 used for gas-testing only, and is useless for lighting purposes. At the last meeting 

 of the British Association the results of an examination of the Ashworth lamp 

 were given by the author ; this lamp burns benzoline, and was found to give good 

 illumination when the wick was raised, and to detect at least 0'5 per cent of fire- 

 damp when the wick was pulled down until it gave a pale-blue flame only. 



In the present paper the author describes a miner's safety-lamp, in which the 

 ordinary fiame can at once be replaced by a hydrogen flame when desired. The 

 use of the hydrogen flame enables the miner to detect readily and with certainty 

 percentages of firedamp, varying between 0'25 and 3-0, and to measure their 

 amount. As soon as the delicate testing is finished the ordinary flame of the lamp 

 is kindled, and can be employed either for illumination, or, if lowered, it can be 

 applied to the detection of percentages of gas larger in amount than those found 

 by the hydrogen flame. The hydrogen gas is carried in a small steel reservoir, 

 slung over the shoulder by a strap, and is introduced through a fine metal tute 

 which passes into the interior of the safety-lamp and terminates near the wick. 

 This composite lamp is at once a good illuminator and an extremely delicate gas- 

 tester. 



Comparative experiments were made with a hydrogen flame, an alcohol flame 

 of the same height, and a small blue benzoline flame, all of which were exposed in 

 air containing 1 per cent, of coal gas. The * cap ' seen over the hydrogen flame 

 was nearly four times as high as that seen over the benzoline flame, and half as 

 high again as that seen over the alcohol flame. 



Many serious accidents have arisen from bringing a ' naked flame ' into spaces 

 in which light petroleum oil had been stored. The vapour of this oil, when mingled 

 with the air in proper proportions, is violently explosive ; and it becomes import- 

 ant therefore to have means of detecting its presence and measuring its amount. 

 The author described tests carried out with the above hydrogen safety-lamp in his 

 test-chamber. They prove that the hydrogen flame can detect one-twentieth of 

 the amount of petroleum vapour which can be kindled in air, and one thirty-sixth 

 of the amount which explodes when mingled with air. 



