1882.] On the Influence of Coal-dust in Colliery Explosions. 441 



Fig. 1. 



I. The average length of fourteen fire-damp flames (12 feet 8 inches) 

 is shown by the dotted line AB.* 



II. The average length of fifteen flames of coal-dnst and pure air 

 (118 feet 6 inches) is shown by the line CD. 



III. The average length of five flames of coal-dust and pure air, 

 augmented by the explosion of small heaps of gunpowder (145 feet), 

 is shown by the line EF. 



All the experiments between the fifteenth and thirty-first are omitted, 

 as they were made while the apparatus was in an imperfect condition. 



In fig. 2 the apparatus is reproduced on a very small scale, so as 

 to show its whole length, as well as the appearances presented by 

 some of the more remarkable flames as they issued from its side, or 

 end. The part which represents the explosion chamber is on the left 

 hand side of the zero line, and the seven sections of the wooden 

 gallery are on its right hand side. The form of the flames was 

 sketched at the instant of their occurrence, with the exception of 

 No. 59, which is correct as to length, and approximately correct as to 

 its other dimensions, but was not sketched until afterwards. The 

 average length of fire-damp flame obtained in the first fifteen experi- 

 ments is shown on each representation of the apparatus by means of 



* After the additional precautions for obtaining a better mixture of the gases, &c, 

 had been introduced, several experiments of this class were made. But as it was 

 found that the fire-damp flame was, if anything, shorter than before, sometimes not 

 exceeding eight or nine feet in length, I did not think it necessary to undertake a 

 new series of experiments, and allowed the original one to remain intact. 



