as the theoretical expressions which, being used daily, assume it,
are, I think, in contradiction with facts (511, &c.). It appears to me to
be as impossible to assume a current of positive or a current of negative
force alone, or of the two at once with any predominance of one over the
other, as it is to give an absolute charge to matter (516. 1169. 1177.).

1628. The establishment of this truth, if, as I think, it be a truth, or on
the other hand the disproof of it, is of the greatest consequence. If, as a
first principle, we can establish, that the centres of the two forces, or
elements of force, never can be separated to any sensible distance, or at
all events not further than the space between two contiguous particles
(1615.), or if we can establish the contrary conclusion, how much more
clear is our view of what lies before us, and how much less embarrassed the
ground over which we have to pass in attaining to it, than if we remain
halting between two opinions! And if, with that feeling, we rigidly test
every experiment which bears upon the point, as far as our prejudices will
let us (1161.), instead of permitting them with a theoretical expression to
pass too easily away, are we not much more likely to attain the real truth,
and from that proceed with safety to what is at present unknown?

1629. I say these things, not, I hope, to advance a particular view, but to
draw the strict attention of those who are able to investigate and judge of
the matter, to what must be a turning point in the theory of electricity;
to a separation of two roads, one only of which can be right: and I hope I
may be allowed to go a little further into the facts which have driven me
to the view I have just given.

1630. When a wire in the voltaic circuit is heated, the temperature
frequently rises first, or most at one end. If this effect were due to any
relation of positive or negative as respects the current, it would be
exceedingly important. I therefore examined several such cases; but when,
keeping the contacts of the wire and its position to neighbouring things
unchanged, I altered the direction of the current, I found that the effect
remained unaltered, showing that it depended, not upon the direction of the
current, but on other circumstances. So there is here no evidence of a
difference between one part of the circuit and another.

1631. The same point, i.e. uniformity in every part, may be illustrated by
what may be considered