26 THE LATE MR. WILLIAM STURGEON ON 



made very particular inquiry about everv^ circumstance that 

 occurred that I considered of importance ; all hands thought 

 the mill was struck, but nobody could find the place. Some 

 thought the whole building was on fire for a moment, and 

 those in the rooms thought the lightning struck every part of 

 the machinery. The fact is, the lightning never entered the 

 building at all ; and it is obvious, from all the circumstances 

 of the case, that the fire originated from a disturbance of the 

 electric element naturally belonging to the machinery ; pro- 

 bably by the influence of the individual flash of lightning 

 already alluded to. Such a disturbance amongst the metallic 

 parts of the machinery would probably give rise to multitudes 

 of sparks, each of which would be sufficient to ignite cotton, 

 wool, or other inflammable material, that happened to be in 

 the way; and, as in this case, the fire originated amongst 

 some loose cotton attached to the carding machine, it is easy 

 to understand how its ignition occurred. I was particularly 

 anxious to ascertain the initial point of ignition, and from the 

 information I received respecting the burning cotton, and the 

 communication of the fire to the floor above, the lower surface 

 of which being that only which had suffered, there appeared 

 no difficulty in understanding where the fire originated. 



In addition to the indications already noticed, I had direct 

 evidence of the existence of sparks by their effects on a roller 

 closely situated to the site of the fire, but not in immediate 

 connection with it. This roller is of wood, about 4 in. long, 

 and l£in. diameter; its hollow axis is filled with lead, and its 

 surface covered with flannel. (I am not aware of the name 

 of this roller, but I send it with the paper for inspection.) 

 The several singed spots on the flannel covering are so many 

 evidences of local fires on the roller's surface, which could not 

 be produced by any ordinary mode of ignition. Nor can it 

 be supposed for a moment that the fire originated from friction, 

 because the machinery had been at rest for more than half an 

 hour ; besides which the parts ignited on this roller are not 



