of Flame and Gases. 405 



I placed the wires of an electrometer, and also of a galvano- 

 meter, in various parts of the affected flame, but could not 

 procure any indications of the evolution of electricity by any 

 action on the instruments. 



I examined the neighbourhood of the axial line as to the 

 existence of any current in the air when there was no flame or 

 heat there, using the visible fumes produced when little pellets 

 of paper dipped in strong solutions of ammonia and muriatic 

 acid were held near each other; and though I found that a 

 stream of such smoke was feebly affected by the magnetic 

 power, yet I was satisfied there was no current or motion in 

 the common air, as such, between the poles. The smoke 

 itself was feebly diamagnetic; due, I believe, to the solid par- 

 ticles in it. 



But when flame or a glowing taper is used, strong currents 

 are, under favourable circumstances, produced in the air. If 

 the flame be between the poles, these currents take their course 

 along the surface of the poles, which they leave at the opposite 

 faces connected by the axial line, and passing parallel to the 

 axial line, impinge on the opposite sides of the flame ; and 

 feeding the flame, they make part of it, and proceed out equa- 

 torially. If the flame be driven asunder by the force of these 

 currents and retreat, the currents follow it; and so, when the 

 flame is forked, the air which is between the poles forms a 

 current which sets from the poles downwards and sideways 

 towards the flame. I do not mean that the air in evety case 

 travels along the surface of the poles or along the axial lines, 

 or even from between the poles ; for in the case of the glowing 

 taper, held half an inch or so beneath the axial line, it is the 

 cool air which is next nearest to the taper, and (generally) 

 between the taper and the axial line, that falls with most force 

 upon it. In fact the movements of the parts of the air and 

 flame are due to a differential action. We shall see presently 

 that the air is diamagnetic as well as flame or hot smoke ; i. e. 

 that both tend, according to the general law which I have ex- 

 pressed in the Experimental Researches (2267, &c.), to move 

 from stronger to weaker places of magnetic force, but that 

 hot air and flame are more so than cold or cooler air : so, when 

 flame and air, or air at different temperatures, exist at the 

 same time within a space under the influence of magnetic 

 forces, differing in intensity of action, the hotter particles will 

 tend to pass from stronger to weaker places of action, to be 

 replaced by the colder particles ; the former therefore will have 

 the effect of being repelled ; and the currents that are set up 

 are produced by this action, combined with the mechanical 

 force or current possessed by the flame in its ordinary relation 

 to the atmosphere. 



