ON THE DEVELOPMENT OF LIGHT FEOM COAL-GAS. 



Ill 



may also result from other causes — such as a sharp bend in the gas supply 

 tube, a speck of dust in one of the orifices of the burner, or, in fact, any- 

 thing that disturbs the even and quiet flow of the gas. One singular 

 example of this is the following : — If a union jet is burning 5 cubic feet 

 of gas at - 5 inch pressure, and a portion of the gas is led away by means 

 of a tube inserted a few inches below the flame, the flame, although 

 diminished in volume, immediately begins to blow. 



In testing flat flames the custom has invariably been to present the 

 flat side to the disc of the photometer ; but although the results so ob- 

 tained are satisfactory in comparing one flat flame with another, they 

 cannot fairly be compared with rattail or Argand flames, which give an 

 equal light all round. The edge of a flat flame gives considerably less 

 light than the side, but the difference between the two depends very much 

 upon the richness of the gas, or, in other words, the opacity of the flame. 

 A flame of gas of low quality is so transparent that an ordinary news- 

 paper can be read through it ; but this cannot be done with a flame of 

 cannel gas except at the lower portion, which in any case offers scarcely 

 any obstruction to the passage s>f light. The following example may be 

 given: — A union jet consuming 5 cubic feet of cannel-gas at '5 inch 

 pressure gave a light of 27 candles when tested in the ordinary manner 

 with the flat side towards the photometer disc, but the edge gave only 

 23 candles, and when rotated so as to give the flame in every position 

 the average result was, as nearly as possible, 26 candles, showing that the 

 ordinary test gave one candle too much, or nearly 4 per cent. In the case 

 of paraffin flat-flame lamps, the difference between the front of the flame 

 and the average all round varies from 4 to 10 per cent. In the latter case 

 the flame is intensely opaque and of a deep yellow colour. All the figures 

 given in this report refer to the flat side of the flame ; and this must be 

 borne in mind in comparing flat with round flames. 



The following table gives the results obtained with Bray's union jets 

 without obstruction to retard the flow of the gas and reduce its pressure ; 

 gas by ordinary test 26 candles : — 



This table gives instructive information as to the effects of mass or 

 quantity of gas, and of pressure. As regards mass, we see that at the 

 same pressure the light afforded by 5 cubic feet of gas per hour varies 

 from 8h to 26 candles, according to the quantity burned ; the lowest 



