94 



On the Detection of Fire- damp. 



[Mar. 31, 



bottle with a little flanged metal nozzle : and this can be attached to 

 an opening in the bottom or side of the oil reservoir of the lamp. 

 The hydrogen, is thus supplied to a copper tube of fine bore which 

 passes through the reservoir and terminates on a level with, and close 

 by, the top of the wick (fig. 1). Before the hydrogen is fed into the 

 lamp, it is gently turned on and allowed to sweep the air out of the 

 flexible tube ; connexion is then made with the lamp, and by careful 

 adjustment of the valve of the bottle the flame is made of the required 

 size. The aperture in the lamp for introducing the hydrogen is 

 closed by an automatic valve when the hydrogen is not in use. 



Several series of measurements of " caps " were made over the 

 hydrogen flame of this lamp in the testing chamber. The flame was 

 adjusted to a height of 10 mm. by viewing it through a metal 

 diaphragm containing a hole 10 mm. in diameter and held outside 

 the lamp ; or by making its tip level with the top of an upright wire 

 fixed inside the lamp and near the burner. A glass cylinder of 

 extra height (90 mm.) 'was fitted into the Ashworth lamp, and a 

 dead -black background was produced by smoking a broad strip of 

 the internal surface of the back of this glass with the flame of a wax 

 taper. The blackened glass not only gave a surface against which pale 

 caps were easily seen, but the dead surface prevented reflections giving 

 perplexing ghosts of the true flame. The percentage of methane in 

 the air in these experiments varied from 0*25 to 3, and the following 

 heights of cap represent the average of a large number of fairly con- 

 cordant readings : — 



With 0*25 per cent, of methane, 17 mm. cap. 

 ,, 0'5 „ 18 



„ I'O „ „ 22 



„ 2-0 „ „ 31 



„ 3-0 „ „ 52 



In the 3 per cent, mixture the tip of the cap disappeared in the 

 opaque metal cylinder of the lamp above the glass. The hydrogen 

 flame therefore became useless for measuring higher percentages, 

 unless it was much reduced in size : but at this point the oil flame is 

 competent to take up the indications with certainty. When very low 

 proportions of gas are to be tested for, the size of the hydrogen 

 flame may be increased with advantage, as is described below ; but no 

 observer could fail to see the smallest cap mentioned above as being 

 produced by the 10 mm. flame. 



The advantage which may be obtained by increasing the size of the 

 hydrogen flame, when small percentages of gas are being looked for, is 

 shown by the results of the following experiments, made by exposing 

 the hydrogen safety- lamp in air containing 1 per cent, and 0*5 per 



