394 EXPERIMENTAL INVESTIGATIONS. 



coil near the orifice or at the commencement of the flame is 

 cherry-red. The galvanometer is deflected to an average of 

 6, the coil near the orifice or at the root of the flame being 

 positive, or related to the farther coil as zinc to platinum in 

 the voltaic trough. On reversing the position of the wires 

 the galvanometer is deflected 6 in the opposite direction. 



This current is not due to thermo-electricity excited in the 

 wires at the point of junction of the platinum and copper, for 

 it is unaltered in direction by a powerful thermo-current ex- 

 cited in these by heating the points of junction by another 

 spirit-lamp. 



By making this thermo-current aid or counteract the flame- 

 current a slight difference in degree is perceptible in the de- 

 flection, according as the point of junction of the one or other 

 wire is heated, but no difference in direction. 



The flame-current, moreover, scarcely affects the short- 

 wired galvanometer, while the thermo-current of copper and 

 platinum whirls the needle to 90. 



It is not a thermo-electric current arising from the unequal 

 heating of the two coils, for it is in the same direction when 

 the farther coil is removed from the full flame, so as to be less 

 heated than the coil at the root of the flame. It is 'also dif- 

 ferent in direction from the thermo-current produced by un- 

 equally heating the coils in similar parts of the flame, or the 

 current described by M. Buff", and to which he ascribes the 

 electricity of flame. The flame-current proper in my experi- 

 ments is opposed to and conquers the thermo-current. 



On advancing the coil from the root of the flame towards 

 the farther coil the deflection lessens, but still preserves its 

 direction until the wires get very closely approximated, when 

 the deflection of the flame-current yields to that of the 

 thermo-current, and the direction of the needle depends on 

 the relative heat of the coils. 



A wire of zinc near the root of the flame with the platinum 

 coil in the full flame gave a much smaller deflection, only 2 ; 

 when the direction of these wires was reversed the current 

 was stronger, the galvanometer marking 5 ; the direction of 

 the deviation was in both instances the same as with platinum 

 and platinum. 



