Chemical Science. 371 
It may, perhaps, be supposed that the effect is due to the power the 
ascending current of vapours has to take positive electricity with it to 
the screen above, and probably the experiment may be considered as 
a proof of the truth of Franklin’s theory, and an argument against 
Dufay. But it is not in a vertical direction only that the aphlogistic 
lamp receives negative electricity from without, and not positive ; but 
in all directions, and from all the concavity of a sphere, of which the 
lamp is the centre. It is not therefore on an emanation in the 
direction of the ascending current of vapours that the effect depends, 
but it resembles rather a radiation like that of light and heat. Se- 
condly, the reciprocal effect essentially requires the actual incan- 
descence of the upper spirals of platina; without this the apparatus 
may be disposed so as to emit a much larger quantity of vapours, but 
in vain. Thus, for instance, 397 grains of platina, which arranged 
properly on a wick would light amadou at two lines distance, and 
keep 400 grains of water boiling, offered nothing like the reciprocal 
action because the upper spires were not ignited ; whilst a spiral only 
a few grains in weight, but incandescent to its extremity, acted in a 
most decided manner. ‘Thirdly, heated iron offers some traces of this 
reciprocity, but only whilst it is in full ignition. The effect cannot, 
therefore, depend on a heated current, which would continue long 
after ignition had ceased. M. Erman has also seen many cases in 
which the iron has had the opposite power to the platina, emitting the 
~ negative and receiving the positive electricity. Finally, as an argu- 
ment against the efficacy of a heated current, undecomposed vapour 
at a high temperature possesses no power,of conducting electricity. 
Without insisting much upon it, M. Erman suggests the following 
explanation of the phenomena: 
There exists two electricities, between which there is a specific difler- 
ence of expansibility: the heat of incandescence acts by augmenting 
this expansibility, in the same manner as the pointed form of con- 
ductors augments the tension, If this augmentation of expansibility 
be very considerable, the specific difference of the two electricities 
disappears in the greatness of the whole effect: this is the case with 
flame ; but there exists a certain degree of heat which augments 
the expansibility in a less degree, and precisely to the point at 
which the most expansible of the two (the positive) is able to over- 
come the constraining force of the circumambient medium, whilst the 
less expansible (the negative) notwithstanding the increment of force 
it has received, has not yet attained to the point at which it can over- 
come the resistance of the medium, ‘The action of the incandescent 
wire is, therefore, according to this view, connected with the pheno- 
mena of the specifically different lights, presented by points positively 
or negatively electrified. Sufficient examination has not been made 
whether points, not incandescent, emit different quantiiies of elec- 
tricity according as they are positive or negative; but the marked 
